The Conundrums of Scaling Up

This post is part of the series of posts where I explore the question “Why does a pilot project succeed, but the following implementations fail?” In my previous post, I looked at how even successful social impact pilot projects fail to show similar results on being replicated due to various factors. I also hinted at pilot projects’ added complexity: while experiments only need to be replicated, pilot projects also need to be scaled up. In this post, I will deal with three essential aspects of scaling up that need to be accounted for: (1) the manner in which we define scaling up and its impact on funding; (2) the locational peculiarities of certain projects and finally, (3) the economics of demand and supply in scaling up.

Scaling up is defined as “expanding, replicating, adapting and sustaining successful policies, programs or projects in geographic space and over time to reach a greater number of people.”

Scaling up may sound simple; however, it is extremely complex and inadequate planning and execution can become the difference between success and failure in the expansion of a social innovation pilot project.

I. Scaling Up and Funding

Understanding Scaling Up

Scaling up can be looked at in terms of being horizontal, vertical or functional in nature.  Horizontal scaling up refers to the expansion of coverage of a project, program, or policy across more people and greater space. Vertical scaling up refers to creating the organizational and political framework needed to roll out on a larger scale. Functional scaling up means going beyond one function (for example, health or education) to include others. The idea of functional scaling up is mostly inapplicable to our area of specific interest; however, in terms of horizontal and vertical scaling up, it is important to balance both of them for the long-term success of any project.

To further understand the nuances of the scaling up framework, let us look at this study of pilot projects. From this study we analyse two different projects which aimed to create electronic records to help address the problem of fragmented health care and incomplete health records for foster youth. The initiative that was developed as part of a state-wide plan, My JumpVault, had an advantage when it came to achieving scale and sustainability. It had broader reach because it was part of a state plan and because it was settled into the state’s political framework. Being a state plan, it also had an easier time accessing potentially significant amounts of federal funding.With state backing, foster care agencies were given access to a core template that was interoperable with state health and child welfare databases, which they could then modify to suit agency practices.

In contrast, Follow My Child was developed and deployed within one foster care agency, which made it difficult to build data linkages within state systems. It may even have been a better system than My JumpVault, but since Follow My Child was not part of the state framework, it could not ensure interoperability between the databases of the relevant organizations and agencies. Ultimately, the project failed to scale state-wide.

So, while reaching more people via horizontal scaling up may be the goal, the same is not possible without creating the organizational and political framework needed via vertical scaling up.Without stable and ongoing funding or significant policy directives that propel a project forward, initiatives often struggle to gain traction.

How and Why Does Scaling Up impact Funding Needs?

Intuitively, the idea of scaling up may seem as simple as increasing the size of a project according to a fixed ratio. The problem is that different stakeholders—including the government, donors, beneficiaries, agencies impacted, and local communities—may have different ideas about what that fixed ratio should be.

For example, the government might want to prioritize scaling up horizontally across wider geographical regions to take a successful policy as far as it can at the lowest cost, whereas donors might want the project scaled up to areas where it has the highest return/impact with respect to the money invested (and they may not be ready to invest money for scarcely populated regions). So, while the government has to look at all its citizens equally, the donors may not feel bound by such principles of non-discrimination and may, instead, be driven by the economics of high returns/impact.

The agencies or organizations involved in the execution of the project may also be enthusiastic about scaling up a successful pilot project across wider geographical reasons. However, first they might want to ensure that vertical capabilities are accounted for (i.e., creating the organizational and political framework required to scale up). Changes in agency practices require changes in policies, administrative procedures, training of employees, etc. While a pilot project may sidestep policy issues by virtue of it targeting only a select group of participants, spreading it widely requires better planning and creation of new frameworks. However, as we shall see below, it is more difficult to find funding for such ‘non-programmatic activities’ which do not directly increase social welfare and rather optimize the project.

 Neither of these scenarios takes into account what might be the voice of the beneficiaries of the said project, who may be asking for something different than the needs of any other stakeholder. For example, they could possibly want higher investment per individual instead of spreading it across beneficiaries.

Therefore, the manner in which a project is defined to be scaled up, which happens depending upon which stakeholder has the strongest voice at the decision-making table (where some of the stakeholders never even reach), is extremely important because it impacts one of the core inputs in a project, i.e., the manner in which funding is to be calculated for the scaling up as it changes the very ratio against which the plan to scale up is created.

How Does Funding impact the Success of a Project?

First we took a look at how the definition of scaling up impacts funding needs. Now we will try to understand why funding is so fundamental to success. It may seem intuitive that inadequate funding is considered one of the three barriers which repeatedly block social innovations from reaching their broadest impact. However, it is still important for us to look at it so as to get a sense of the scope and magnitude of this problem and why it is encountered even for projects which have clearly shown positive results at their pilot stage. A study notes that the “failure to replicate innovative social programs is usually attributed to problems of strategy and management (whereas) much of the time, it is simply a problem of money.”

Raising funds for scaling up is generally more difficult than getting funds for a pilot project. While funders are generally drawn to early stages of social innovation projects, preparing to scale increases organizational costs because investments are needed to upgrade technology, hire senior-level talent, and improve infrastructure. In addition to that, each site needs capital to develop and adapt the model. It is not just the increase in magnitude of funding required for scaling up that makes it an issue. It is also that scaling up is not an activity which produces immediate results as pilot projects generally do. Therefore it can seems less appealing to funders as they struggle to align their expectations with the long-term nature of social innovation.

Further, as the network of such projects grows, they generally also need funding for a national office, which normally performs ‘non-programmatic activities.’ As these ‘activities’ do not seem to have a direct impact of increasing social welfare, they are difficult to secure funding for. The importance of such ‘activities’ is rarely taken into account. Having a strong central office is extremely important to ensure support and management discipline that enables the on-ground programs to be executed more powerfully.

Establishing a reliable source of funding as well as a standardized flow of money increases the odds of success for two reasons.

Firstly, the leaders can direct their time and energy towards the program instead of finding ways to raise funds. Secondly, it can help to minimize the pressures created by funders’ varying interests.

For example, we can look at the case of Jumpstart, a program which matches young children who are struggling in preschool with college students (called corps members) for a one-year relationship. These corps members have a program guide specifying how to develop a customized curriculum for each child and offering a range of reading activities associated with each developmental need. However, some funders were interested primarily in engaging older youth in the community, while others focused on literacy (and thus the preschool tutees). The cumulative effect of such pressures can lead to loss of focus in the program which may end up in diminished results.

II. Location and Scaling up

Pilot projects are generally started with the aim of being scaled up in a particular context and its effectiveness is often context dependent. But when a certain social innovation project shows positive results for a problem faced across the board, it is often attempted to be scaled to the state level, or sometimes, even on to the national level. In this process, it is generally forgotten that while a problem in the new context may be similar, the solution may not be. It is not possible to simply replicate an innovation from one location to another. It is extremely important to understand the specific context that comes with each site. It will have its own culture, existing infrastructure, policies, political dynamics, demographics, socioeconomic statuses, geography; all of which can tremendously impact the success of a project.

For example, Summer Bridge is an intensive summer program which is meant to prepare talented youngsters from diverse backgrounds to succeed in college. It was initially hosted in independent schools, and when the organization tried to replicate it in public schools, the model proved difficult to implement. Board members debated the causes and consequences; some worried that the public schools’ bureaucracy and thin resources would constrain the program, while others believed the independent schools’ attractive campuses were part of the program’s allure. Effective replication often depends on holding constant or ‘standardizing’ the context within which a program will operate. However, in social innovation projects, as I discuss below, only limited control can be exercised over such ‘context’.

The hugely popular Harlem Children’s Zone (HCZ) is a recent example of a case where context was key. The model has come to be known as the cradle to college pipeline, or conveyor-belt strategy, and it has been very successful in helping children in Harlem get a proper education and go on to college. The success of this model has led to other districts attempting to fix education in their high poverty areas by attempting to replicate HCZ. These programs have had varying degrees of success. In a study of two different districts where the model was adapted, it was found “that different cities will never have the same experience due to existing challenges, varying funding and community support.” Another attempt in Durham, North Carolina,  to replicate the model was criticized for having failed to understand the historical and cultural identity of the borough and implementing the model built in Harlem’s context without customizing it for the special needs of Durham. HCZ’s Practitioners Institute, which was created to “share information about our work, teach best practices, and provide pivotal guidance to groups from across the U.S. and around the world invested in replicating our model” also notes the importance of not repeating but customizing. It states and rightly so that “each community has unique needs to be addressed and resources on which to capitalize in building a pipeline to serve local children and families at scale.”

A simple project, such as recovering “edible but not sellable” fresh fruits and vegetables and swiftly distributing these nutritious foods to low-income people via food banks, pantries, and other distribution services—a pilot which one would expect to be easily replicable—turned out to be “anything but fast and obvious.” This extensive study of ‘replicating’ produce programs over 19 years shows that customization and innovation for every site individually is the key to successful growth of a project in different locations. The aim to “to produce an identical version, repeatedly, as in commercial franchising, does not fit social innovation projects. While commercial franchisees simply set up a shop for customers whose preferences are generally the same, social innovation projects look to change a status quo which varies from place to place. Therefore, it is extremely important that projects account for local conditions and customize as per the barriers at the new sites. Applying solutions built in another place to a new site is much more complex than simply copying them and requires nuance to understand the intricacies and peculiarities of the new site.

The question, however, remains that if customizing, not repeating is the key to success, then what is the core portion and what needs to be customized? The answer simply varies from project to project. The status quo of the places where the project is being replicated is different. Projects need to account for different conditions for a replication to be successful.

III. Economics of Scalable Projects

Another factor that needs to be accounted for is the viability of scaling up a project. A project may be aimed towards providing ‘something’ which is valuable because it is scarce. If such projects are scaled up and the ‘something’ derived value only because of its scarcity, it would lose such value and the project would fail. This is especially applicable to social innovation projects which aim to create livelihood or employment opportunities through grants or subsidies as we shall see in an example below. A similar issue arises when the outcome targets for a programme are designed relative to a larger population as explained later.

For example, No Lean Season is an innovative program that wascreated to help poor families in rural Bangladesh during the period between planting and harvesting (typically September to November). During this season, jobs are generally not available and a lot of rural poor suffer due to seasonal poverty. No Lean Season aimed to solve that by giving small subsidies to workers so they could migrate to urban areas for the season. In small trials, it worked great. However, when it was scaled up to 699 villages from 82, the results of an intensive monitoring study found that “the program did not have the desired impact on inducing migration, and consequently did not increase income or consumption.” One of the reasons for the decline in success was that, “when delivered as a scaled program, providing subsidies for seasonal migration ceases to have the effects observed when delivered at smaller scale.” As the program is scaled up individuals may be less likely to migrate if they perceive that hundreds of other people who would compete with them for the same jobs are also migrating.

In terms of designing targets, while relative measures may work for pilot programs, they are inappropriate for programs which are to be scaled up. A simple example might help us understand this phenomenon in a better manner: for a county pilot program which seeks to improve the quality of education in the area, it may make sense to measure achievement of targeted students relative to the state percentile. That is, the program may attempt to move more students to, say, the top half of the overall state population. However, if the program is deemed to be successful and is to be scaled up to all schools in the state, the targets need to be redesigned and cannot obviously be in percentile terms.

Therefore, it is important to ensure that a project which is scaled up is not designed to provide resources which derive their value from being scarce, such that scaling up will lead to their value being diminished. Similarly, its targets should be ideally designed in non-relative terms.

In this post, under the first theme we tried to understand the different ways in which scaling up is defined namely vertical, horizontal and functional. We then looked at how the various stakeholders differ in their understanding of the ideal manner for scaling up of pilot projects. Under this theme we also looked at the role of funding in scaling up a pilot project successfully. Under the second theme of location and scaling up we noted that when a project shows positive results for a problem faced across the board, it is often attempted to be scaled beyond the original context it was meant for. In such cases, we tried to understand the importance of contextualizing a pilot project before replicating it new areas. Finally while analyzing the third theme concerning the economics of scaling up, we looked at certain kinds of projects which lose their viability on being scaled up. Thus, in this post we dealt with three specific issues pertaining to the scaling up aspect of replicating a pilot project while we had dealt with the general issues in replicating a project in my previous post.