- EI blog - http://blog.ei-india.com -

Demystifying Impact Bonds

The New Education Policy (NEP) talks about increasing funding to education sector to 6% of GDP. Apart from additional financing, there is a need to ensure that existing financing is better spent and reaches the most vulnerable groups of the society effectively; i.e., each rupee spent on a child should deliver a higher social return on investment. Specifically, in the Indian education landscape, several reports (including those from World Bank, UNICEF, ASER, etc) highlight that the school access for students have not necessarily translated into the desired levels of learning; i.e. paying for inputs does not necessarily mean impact.

Given the financing gap faced in achieving their Sustainable Development Goals (SDGs), governments and donors in many countries have increasingly adopted Results Based Financing (RBF) approaches to increase the effectiveness of their development spends (World Bank, 2019)[1]. In light of that, a relatively new model of financing, the Development Impact Bond (DIB) makes a case to ensure better outcomes for the beneficiaries. This article provides an introduction to the concept of RBF and uses the example of DIBs (or SIBs[2]) with that. It then describes the different players that come together to make a DIB happen, an overview of the global and Indian landscape, and finally a critique of the benefits and limitations of DIBs.

 Introduction to Impact Bonds and Results-based financing:

First launched in 2011 by Social Finance UK, impact bonds fall under the broader category of Results-Based Financing (RBF) approaches. RBF refers to any program that links funding to achievement of pre-agreed results. Historically, there are many inputs based metrics that are tracked rigorously by governments and donors (eg. number of schools, classrooms, teachers) and some outputs (eg. # of teachers undergoing training, marks scored by students). Over the years, different RBF approaches have emerged in an attempt to increase the effectiveness of traditional input-based funding models. Two key factors that distinguish these different RBF approaches are: (a) How much of the overall payment is linked to outcomes? and (b) Who bears the financial and implementation risk of outcomes not being achieved?

In Results-based aid instruments such as the Program for Results of the World Bank, the Government agencies usually bear the implementation risk and receive partial payouts on achievement of outcomes from the donor (the World Bank). In performance-based contracts, the service providers bear the implementation risk and receive a part of their payments from governments only on achieving outcomes. However, governments and service providers, due to various operational and financial limitations are not always the best party to assume this implementation risk. Impact bonds, in such cases, provide a mechanism to transfer this risk (fully or partially) to investors who provide upfront finance with the intention to generate positive, measurable social and environmental impact alongside a financial return. Note that Results-based aid often cannot be clearly distinguished from results-based finance. Therefore, there are a number of hybrid schemes that combine characteristics of both approaches.

An impact bond is a results-based contract whereby outcomes funders, service providers, and investors are bound by a common goal—the achievement of outcomes. Impact investors provide upfront capital and support service providers to reach the targeted outcomes. If the targets are met or exceeded, government (in the case of a Social Impact Bond) or philanthropic donors (in the case of a Development Impact Bond) pay back the investors their principal plus a return. Independent evaluators measure the success of the program in reaching the target learning outcomes, triggering payments. In general, the more successful the programme, the greater the return to investors, usually up to some cap.

 Key actors and roles:

  1. A service provider who delivers social services to the targeted beneficiary segment. The service providers can be either non-profits/ NGOs or for-profit private organisations.
  2. A risk investor provides working capital to the service provider and bears the risk if the service provider does not achieve outcomes.
  3. An outcomes funder repays the investment on the achievement of outcomes
  4. An independent evaluator assesses the impact and level of outcome achieved
  5. An intermediary who designs and structures the impact bond, convening all stakeholders, raising capital and structuring and managing the deal between all parties

Figure 1 : Social Finance India

Impact Bonds Landscape

Globally, Impact Bonds are known by different names: as Social Impact Bonds (SIBs) in the UK, as Pay for Success (PFS) contracts in the US, and as Social Benefit Bonds in Australia. Based on Brookings Global Impact Bonds Database, a total of 194 impact bonds have been contracted across 33 countries in six key sectors (image below). 22 impact bonds have been contracted in education globally – most of them focusing on early childhood interventions and academic support for out of school children.

Figure 2 Source: Brookings Institution Global Impact Bond Database, July 2020


These 194 bonds represent commitments of $421 million in investment and $460 million in outcome funding[3]. More than 90% of these bonds are Social Impact Bonds with the government as the Outcome Funder. The United Kingdom, the United States, the Netherlands, Australia, and Portugal are the largest markets for impact bonds. Only 17 impact bonds have been contracted in low- and middle countries, mobilizing around $48 million of investment. India leads the developing countries with the most contracted DIBs (Annex 3 provides a snapshot of the education DIBs in India). There is also a huge variation in the average number of beneficiaries supported in impact bonds across countries, largely because of different social context and specific interventions addressed by these bonds. For instance, with just 3 out of the 194 impact bonds listed in Brookings database, India accounts for 80% of the total impact bond beneficiaries globally[4]. While the average number of beneficiaries targeted by an impact bond is just under 12,000 (influenced by the few large outlier impact bonds), the median is just 500.

Characteristics of DIBs:

  1. Rigorous focus on outcomes and beneficiaries along with adaptability for Service Providers: DIBs tend to keep a laser sharp focus on the outcomes that were originally agreed on. Hence they provide clear direction on what every member of the implementing organisation needs to focus on. Intermediate evaluations are planned to identify the gaps to target and hence spur innovation for the service providers to make changes to achieve the goal. For instance, in the course of the Educate Girls DIB, the first and second years showed lagging growth against the pre-set enrolment and learning targets[5]. However, since success was tied to outcomes and not activities or inputs, Educate Girls was able to tweak and adapt its classroom management techniques, curriculum and content, nature of training and assessments as well as the community mobilization process. Frontline staff were also empowered to make decisions and shift resources to areas that needed more attention than others. The DIB thus provided room for the service provider to experiment with innovations in its program, and do what was necessary to ensure the ultimate outcomes were delivered and surpassed.                                                  
  2. Scaling evidence-based solutions is another key advantage of opting for an impact bond: Non-profit organizations working at the grass-root level may sometimes have operations in several states, but may struggle to roll out state-wide, systemic programs. The impact bond model can help them scale their interventions using the prior evidence to build investor confidence and support the price discovery for outcomes. In the long run, the rigorous data and evidence generated from these initial impact bonds can help the governments to launch intervention(s) at scale.                                                    
  3. DIBs incentivize collaboration: DIBs by design involve multiple stakeholders with different competencies in co-creation from the beginning thereby enabling cross-sector partnerships. Moreover, they also encourage collaboration between Service Providers with diverse strengths as the targeted outcomes are clearly defined upfront. E.g. in the case of the QEI-DIB – Educational Initiatives, a for-profit organisation, which makes the Mindspark learning software partnered with Pratham Infotech Foundation, a not-for-profit organisation with significant on-ground experience of deploying EdTech in government schools. As with most partnerships, this also started on a rocky ground of two organisations with very different cultures, types of talent, styles of working coming together. However, because the DIB was very clear in terms of what the organisations were expected to achieve in terms of outcomes (learning improvement in mathematics and language) – the two partners agreed on the inputs (deploying the hardware in schools on time, getting government permissions to start immediately and staffing and training a team of 30+ people on the ground) and outputs (120 mins of usage of the software per child) that would help them achieve the outcomes target. In the end, the consortium exceeded learning outcome targets by a large margin and making the first year a success.                                                                                                                                                                                                    
  4. DIBs may have additional costs: One of the commonly cited limitations of impact bonds is that they are more expensive for the outcome funders (whether philanthropic agencies or the government) on a per beneficiary cost compared to traditional grant-making because of the added costs of independent evaluation, administrative costs of structuring the bond and setting up performance management systems, as well as the return paid to investors. Let us briefly look at each of these additional cost elements to assess what value they add. Firstly, the rigorous and independent evaluation adds significant value by increasing transparency of how the donor money was spent and whether it achieved the desired outcomes. In fact, we believe robust program evaluation should be considered as an important component in all donor funded programs, whether grants or DIBs. Secondly, the experience of the two Indian education DIBs has shown the value of investing in performance monitoring and well-designed bond structures. For example, as mentioned earlier, the data systems set up in the Educate Girls DIB helped the frontline staff to diagnose the key issues and accordingly plan the right set of course-corrections needed to achieve the outcomes. Lastly, the return paid to investors is the premium that outcome funders are paying for transferring the risk of intervention not achieving outcomes as well as time value of money.

A like-for-like comparison of costs in a DIB vs. the traditional funded models without any risk transfer, should therefore, include an average cost of delivering outcomes of both successful and unsuccessful programmes.  Moreover, DIBs are in a nascent stage across different social sectors and the overall costs will reduce with economies of scale as more bonds are launched, and evaluation methodologies and contracting structures get standardized.

  1. DIBs disincentivise innovation: It is often argued that impact bonds do not incentivise real innovation as investors prefer to support projects with a proven track record, assuring their returns. It is true that DIBs are well suited, and thus often used, to scale up evidence-based interventions and are not the most suitable option for funding promising but unproven interventions. For instance, DIBs may not be a suitable solution for funding the development of new innovative ed-tech content and delivery platform, which may be better funded through a grant or venture capital investment. Having said that, several case examples of DIBs show that even well proven interventions require great innovation in service delivery when they are scaled up manifold and replicated in different geographic or beneficiary settings.

In summary, DIBs should not be seen as a panacea for all development sector challenges and are more suited in certain contexts such as for scaling evidence-based solutions and in cases where greater service provider capacity to adapt/innovate adds significant value to the intervention. Similarly, DIBs are more suitable for funding preventive interventions such as reducing the risk of Type-2 diabetes by implementing behavioural and lifestyle changes but will not be an ideal option for say disaster relief efforts that require immediate reactive approaches.

Additional resources to understand more about impact bonds:

  1. A brief by Brookings Education on the size and scope of the development impact bonds market :https://www.brookings.edu/wp-content/uploads/2020/09/Impact_Bonds-Brief_1-FINAL.pdf
  2. About the Educate Girls Development Impact Bond:
  3. About the Quality Education India DIB: https://qualityeducationindiadib.com/
  4. About Haryana Early Literacy DIB: https://socialfinance.org.in/2020/07/20/government-of-haryanas-landmark-pay-for-success-program/
  5. Know more about how targets are set and outcomes are evaluated in a DIB: https://blog.ei-india.com/2020/08/31/development-impact-bonds-target-setting-and-outcome-evaluation/
  6. More about Paying for Outcomes at Scale in India: https://www.brookings.edu/wp-content/uploads/2019/11/Paying-for-education-outcomes-at-scale-in-India-FINAL-FOR-WEB.pdf

Appendix – Snapshots of education DIBs in India

  1. Educate Girls DIB
Outcome metrics Student enrollment and learning outcomes (improvement in Hindi, math and English levels on the ASER test, relative to the control group)
Beneficiaries Students in Grades 3-5 in 66 treatment schools (7,318 students were in Grades 1-5 in treatment schools at baseline)
Upfront capital committed $270,000
Maximum outcome


Investor returns IRR of 15%
Timeline Contracted 2015

Ended 2018

Key milestones One outcome payment disbursed in final year
Validation methodology Pre-post test using validated administrative data (enrollment) Randomized controlled trial (learning outcomes)
Service providers Educate Girls
Outcome funders Children’s Investment Fund Foundation
Investor UBS Optimus Foundation
Intermediary Instiglio
Evaluator IDinsight

2. Quality India Education DIB

Outcome metrics Learning outcomes: improvement in literacy and numeracy outcomes relative to comparison group


Payment metric: number of beneficiaries multiplied by improvement in learning

Beneficiaries targeted 200,000 students in Grades 1-8
Upfront capital committed $3 million upfront, with capital recycled each year if outcome metrics are achieved
Maximum outcome


$9.2 million over 4 years
Investor returns IRR of 8% if targets met
Timeline Start of services: April-June 2018 Contracted: August 201815

Yearly results released: July 2019, 2020, 2021 Final results: July 2022

Key milestones Performance reviewed annually, with outcome payments disbursed each year
Evaluation methodology Quasi-experimental design—learning improvements measured relative to a comparison group
Service providers (see Table 4 for further details) Gyan Shala SARD


Educational Initiatives (Mindspark)/Pratham Infotech Foundation*

Outcome funders Michael and Susan Dell Foundation (MSDF) and consortium of funders convened by the British Asian Trust (Tata Trusts, Comic Relief, Mittal Foundation, British Telecom)
Investor UBS Optimus Foundation
Intermediary Dalberg
Evaluator Gray Matters India

3. Haryana Early Literacy DIB

Beneficiaries ~115,000 Grades 1 and 2 students, and ~7,500 teachers and teacher educators across 7 districts (~3,300 schools) of Haryana
Outcome Funding committed ~ $2.3M
Max. Performance Guarantee ~ $0.45M (20% of Outcome Funding)
Timeline Contracted 2019
Ends in 2022
Outcome Metrics Improvement in reading ability of Grade 1 and 2 students to understand appropriate texts (based on the Early Grade Reading Assessment (EGRA) tool)
Validation methodology Quasi-experimental difference-in-difference approach with control and treatment groups to estimate intervention’s impact
Service Provider Language and Learning Foundation (LLF)
CSR Outcome Funders IndusInd Bank, SBI Foundation
Performance Guarantor Central Square Foundation
Intermediary Social Finance India
Outcomes Evaluator Educational Initiatives

[1] World Bank, 2019: https://www.worldbank.org/en/news/feature/2019/06/28/banking-on-impact-what-you-need-to-know-about-results-based-financing

[2] In case the outcome funder is a government then it’s called a social impact bond (SIB) and behaves similarly. For the purpose of this article a reference to DIB can be to either a DIB or SIB.

[3] https://www.brookings.edu/research/what-is-the-size-and-scope-of-the-impact-bonds-market/

[4] Are impact bonds reaching the intended populations? Brookings, Global Economy and Development

[5] https://www.idinsight.org/reports-2/2018/8/22/evaluating-the-educate-girls-development-impact-bond