Measuring Impact
Measuring Impact
Measuring Impact
“impact”
in impact investing
Executive Summary 5
1. Context 11
5. Cross-Cutting Themes 50
5.1 Incentives 50
5.2 Additionality 53
7. Conclusion 57
List of Figures
Figure 0.1 Continuous Cycle of Measurement Objectives
Figure 0.2 Map of Measurement Methodologies to Measurement Objectives
Figure 0.3 Summary of Measurement Method Analysis
Figure 0.4 Integrated Model of Impact Measurement
Impact investors employ a number of methods to pur- 3. Mission alignment methods measure the execution
sue the objectives outlined above. By identifying pat- of strategy against mission and end goals over time;
terns that we found in our research, we have categorized examples include social value criteria and scorecards
four impact measurement methods: used to monitor and manage key performance met-
1. Expected return takes into account the anticipated rics.
social benefits of an investment against its costs, dis- 4. Experimental & quasi-experimental methods are after-
counted to the value of today’s value. This expected re- the-fact evaluations that use a randomized control trial
turn metric can take various forms; examples include or other counterfactual to determine the impact of the
Social Return on Investment (SROI), Benefit Cost Ra- intervention compared to the status quo.
tio (BCR), and Economic Rate of Return (ERR).
2. Theory of change and logic model explain the process Impact measurement methods generally serve specific
of intended social impact. Specifically, logic model is objectives in the investment cycle. Mapping the methods
a common tool used to map a theory of change of an against the objectives provides a view of how each of these
organization, intervention, or program by outlining methods can accomplish the different objectives. The fol-
the linkage from input, to activities, to output, to out- lowing graphic illustrates how such methods have been
comes, and ultimately to impact. applied to the various phases of impact measurement.
Expected Return
• SROI
Theory of Change
• Logic Model
Quasi-Experimental &
Experimental Methods
• RCT
• Historical baseline
• Pre/post test
• Regression discontinuity design
• Difference in differences
Expected Return • To estimate expected social • Can provide a disciplined • May unfairly penalize in-
return in assessing potential approach for decision making terventions working with the
investments • Offers opportunity for orga- most challenging problems
• To monitor and evaluate nization to speak a common and populations
the social performance of language • Can be perceived as inexact
investments • Similarity with return on and constantly changing
investment can help gain • Expected return calculations
private sector trust are only as strong as the data
that feeds them
• Risk of temptation in using
expected return figure as
standalone metric for funding
decisions
• Not applicable to interven-
tions without quantifiable
benefits
• Does not take into account
catalytic effects
Theory of change • To understand path to in- • Provides an easy to un- • Identifying indicators to
and logic model tended impact as part of due derstand framework that is assess outcomes can be
diligence familiar in the social sector challenging
• To provide a framework for • Is a versatile tool that can • Lends itself to risk of reduc-
goal setting serve multiple purposes ing social change to a linear
• To track and monitor prog- • Allows investors to overlay process
ress of investment dimensions that are import-
• To provide targets for incen- ant to mission
tive schemes • Allows investors to identify
• To provide a framework for underlying impact assump-
illustrating impact logic in tions for further review as
reporting necessary
Mission align- • To monitor impact investor’s • Surveys and screens are • Survey results or scorecards
ment methods portfolio against its mission inexpensive, straightforward are only as meaningful as the
• To monitor impact of invest- ways to monitor mission data collection methods or
ee against its mission alignment KPI metrics that they capture
• Scorecards may resonate • Scorecards may not allow
with investors due to familiar- for direct comparisons across
ity with balanced scorecard in different investments
business
8 measuring the “impact” in impact investing
Experimental & • To assess outcome for pay- • Experimental methods allow • Experimental methods can
quasi-experimen- ments in Social Impact Bonds for robust cause-and-effect be expensive and resource
tal methods and other impact investments attribution intensive
• To test hypothesis of an • Quasi-experimental methods • Experimental methods not
investor’s theory of change can provide some attribution suitable in many situations,
• To assess impact risk of a evidence with more flexibility e.g. environments that cannot
potential investment and lowest cost be controlled, interventions
• Both of these methods can that are insufficient to drive
help to demonstrate addition- outcomes on their own,
ality of impact situations where randomizing
beneficiaries may be unethical
• Quasi-experimental methods
may be limited in their ability
to rule out exogenous factors
• Drawing upon existing experimental or quasi-exper- We encourage impact investors to map their own theory
imental studies to test the hypothesis underlying the of change to understand how their investments translate
causal links into intended impact, and to conduct necessary research
or evaluations to validate assumption.
After due diligence, the investor works with the investee
in: The simpler adaptation
Our framework also proposes a simpler version for those
• Determining key performance indicators (KPIs) to
that are just starting out. As a first step, we encourage
track on the monitoring scorecard.
investors to work with entrepreneurs to develop a logic
• Gathering and analyzing data on the KPIs post invest- model to map their venture’s theory of change so the
ment to monitor the social impact performance of the investor can understand and evaluate its path to impact.
investee. In the pre-approval stage, we suggest adopting social
• If required, using a quasi-experimental method evalua- value criteria to rate investments, and to monitor the
tion in the evaluation stage. investee’s progress post-investment.
Note: Investee maturity should be determined by the impact investor based on the investee’s size,
reach, budget, or years in existence
10 measuring the “impact” in impact investing
1. Context
Since the coining of the term in 2007, impact investing • Enterprises or Investees can use metrics to determine
has captured the interest and imagination of the busi- progress and improve their impact.
ness world, governments, and social sector organiza- • Beneficiaries can participate to help improve the effec-
tions alike. In 2009, the Monitor Institute estimated the tiveness of social or environmental gains.
size of the impact investing market to be $500B over the
next decade1, with some analysts believing that this is a
In addition to our own learning, our intention is to be
conservative estimate. Impact investing has taken on a
helpful to impact investing practitioners by examining
global footprint and has gained the attention of govern-
how other impact investors measure the impact of their
ments around the world, as evidenced by the creation
investments, which may be useful as they build upon
of the G8 Impact Investing taskforce and the White
their existing impact measurement methodologies.
House’s Social Innovation Fund.
Further, it is our hope that the findings of this project
will be useful for a few other stakeholders:
This growth has produced an unprecedented focus on
measuring the social impact resulting from these invest-
• Social sector organizations (e.g. nonprofit service deliv-
ments. Prominent literature and research has emerged
ery providers, social enterprises) interested in learn-
in recent years on the topic of impact and performance
ing about how impact investors are currently thinking
measurement;2 however, the impact measurement space
about measuring impact; this can help them in struc-
is nascent and largely unstructured, with a number of
turing their own monitoring and evaluation functions.
methodologies and players emerging in a seemingly
uncoordinated fashion. Our project builds upon the foun- • Traditional funders (e.g., foundations, aid agencies)
dation laid by recent publications. It takes a more tactical interested in various ways of impact measurement
approach to deepen understanding of specific practices taken on by impact investors, and how they may learn
and methodologies investors are using to measure the from these methodologies as they look to measure the
social impact generated by their investments. impact of their own work. This may also be particular-
ly useful for foundations that are considering involve-
As illustrated by Reeder & Colantonio in their 2013 paper3, ment with impact investing, as this helps increase their
impact measurements can in theory fulfill a number of familiarity with emerging players and landscape. It
critical tasks benefiting different stakeholders: may also shed light on potential areas to fund as they
are looking to build an enabling ecosystem to tackle
• Investors can monitor progress and find out the extent
social challenges.
to which their actions are helping or hindering social
goals. • Evaluators (e.g. impact measurement organizations) to
increase the adoption of measurement and evaluation
• Portfolio managers can use impact measurements to
methodologies, and to look beyond their own work to
estimate and select their investments and benchmark
identify potential areas for collaboration.
the effectiveness of different investments.
1 Freireich, Jessica and Katherine Fulton. “Investing for Social & Environmental Impact.” Monitor Institute. January 2009.
2 For example, the G8 Social Impact Investment Taskforce published a report on “Measuring Impact: a subject paper of the Impact Measurement
Working Group” in 2014 In 2013, London School of Economics (LSE) Cities partnered with the Young Foundation to publish a series of reports on
the topic, including “Measuring impact and non-financial returns in impact investing: A critical overview of concepts and practice.” The European
Venture Philanthropy Association Knowledge Centre also published “A Practical Guide to Impact Measurement” in November 2012.
3 Reeder, Neil and Andrea Colantonio. “Measuring Impact and Non-financial Returns in Impact Investing: A Critical Overview of Concepts and
Practice” EIBURS Working Paper 2013/01. London School of Economics and Political Science. October 2013.
12 measuring the “impact” in impact investing
For the purposes of this paper, we define impact as orga- In addition to reviewing prominent literature on this
nizations (i.e. not individuals) that make financial invest- topic we conducted 20+ interviews with employees across
ments with the intention of generating a social and envi- a wide range of domestic and international organizations
ronmental impact alongside a financial return. We look in the private, social and public sectors. Interviewees
at investments in both domestic (US, Canada, and UK) included practitioners from large banks, for-profit and
markets and in developing countries. We define “impact” non-profit impact investors, venture philanthropists,
not as the furthest step of a logic model as is depicted in foundations, social impact bond intermediaries, govern-
some literature, but in a broad sense often used by prac- ment agencies, market-building organizations, academ-
titioners to describe the positive change generated by an ic institutions, and other sector experts.
investment. Overall, we define impact measurement as
the activities the investor organization takes to evaluate The following table outlines the organizations that we
and report on the social change generated. studied as part of this project.
13 measuring the “impact” in impact investing
• Social Innovation Fund • New Profit • RSF Social Finance • LGT Venture Philanthropy
• Robin Hood Foundation • Draper Richards Kaplan • Bridges Ventures • Omidyar Network
• REDF • Grassroots Business Fund • Bank of America Merrill Lynch
• Acumen • JP Morgan
• Root Capital • Goldman Sachs
Other interviewees
Evaluators • MDRC
Consultants • FSG
We recognize that this is only a small sampling of a however, our hope is that our study highlights some
large and rapidly growing industry, and that our findings prominent methodologies and examples to encourage
may not necessarily be representative of the entire field; learning and discussion.
14 measuring the “impact” in impact investing
Expected Return
• SROI
Theory of Change
• Logic Model
Quasi-Experimental &
Experimental Methods
• RCT
• Historical baseline
• Pre/post test
• Regression discontinuity design
• Difference in differences
4 Ebrahim, Alnoor, and V. Kasturi Rangan. “What Impact? A Framework for Measuring the Scale & Scope of Social Performance.” California
Management Review 56, no. 3 (Spring 2014): 118–141.
16 measuring the “impact” in impact investing
It can take the form of a %ROI, a ratio, or a Net Present • Assign monetary values to those outcomes considered
Value (NPV) number. significant
• Reassess significance of the outcomes in light of the
There are two types of SROI. The first is evaluative, relative values
which is conducted retrospectively and based on actual • Calculate the SROI ratio (impacts/inputs) for these out-
comes
outcomes that have already taken place. The second is
forecast, which predicts how much social value will be
created if the activities meet their intended outcomes.
This is especially useful in the planning stages of an
activity, or if existing data does not enable you to compute
Manage value
an evaluative SROI.
5 Nicholls, Jeremy, Eilis Lawlor, Eva Neitzert, and Tim Godspeed. A Guide to Social Return on Investment January 2012.
6 GIIRS & SROI Network. “GIIRS and SROI: What is the relationship?” February 2013.
17 measuring the “impact” in impact investing
In this methodology, the outcomes are ideally deter- Beyond its use in due diligence processes to make
mined through a process that involves those experienc- funding decisions, organizations use expected return for
ing the outcomes. Negative (including unintended ones) several other purposes7:
and positive outcomes should both be included. • Monitoring and reporting: Some investors continue to
adjust expected return calculations as more informa-
Methodologies that apply a similar approach include tion becomes available throughout the life of an invest-
Benefit Cost Ratio used by Robin Hood Foundation, and ment. This can inform reports to stakeholders on how
Economic Rate of Return as used by Grassroots Busi- much social value has been generated over the life of
ness Fund. These variations on expected return will be an investment.
explored in more detail in Spotlights under the Applica-
• Vocabulary and communication: By translating dispa-
tion section below.
rate output and outcome metrics into economic terms,
expected return measurement enables grant-mak-
Application ers and investors to speak a common language and
Generally, we discovered that the expected return assess- compare investments directly – both upfront and after
ment is used as an anticipatory means of evaluating the fact.
which investments the organization would like to fund • Transparency: Expected return measurements allow
rather than a retroactive measurement of outcomes grant-makers and investors to remove personal biases
and impact of investments already made. Specifically, in investment decisions. While expected return cal-
funders use measures of expected returns to internally culations alone are certainly not the reason to make
rank potential grant applicants, comparing the impact of investment decisions, they enable everyone to see
similar and dissimilar programs in a common language, project financial implications of investments in a trans-
and to assess a potential investment’s fit. parent way.
• Baseline for investment: The expected return calcula-
As in the private sector, expected return is certainly not a tions hold the impact investor responsible for a metric,
guaranteed rate of return. Nevertheless, it is useful in fore- incorporating into the social sector the accountability
casting the future value of an investment or an entire port- that the private sector enjoys.
folio, and provides a benchmark from which to compare.
REDF funds social enterprises that they believe (1) address this market failure in employment, and (2) have a
sustainable, long-term business model long term. With SROI, REDF saw an opportunity to measure and bring
together both of these missions by blending social cost savings and financial analyses. This resulted in “a snap-
shot that [REDF calls] the “blended value” – the financial and social return achieved by these social enterprises.”8
When using SROI to make investment decisions, REDF Economic Social Blended Value SROI
computes the increase in economic value created as a result Value
+ Value =
or Total Business and
of the investment, as well as the increase in socio-economic Social Benefits
or social value generated by the investment.
7 Weinstein, Michael M. and Cynthia Esposito Lamy. “Measuring Success: How Robin Hood Estimates the Impact of Grants”. Robin Hood.
February 2009.
8 Javits, Carla I. “REDF’s Current Approach to SROI.” REDF. May 2008. Pg. 1.
18 measuring the “impact” in impact investing
For instance, if REDF wants to provide a down payment for purchasing a building that will house a social
enterprise in San Francisco, they estimate first economic value over time: estimated net income of business
performance attributed to investment. They then determine socio-economic or social value over time; in this
case, these may include federal taxes from new jobs created by the enterprise, and government savings from
food stamps and Temporary Assistance for Needy Families (TANF), minus the estimated social costs. Adding
these together gives the total benefits, from with a net present value can be determined and an SROI estimated.
A 2000 REDF report9 on SROI provides a simple illustration of two cash flows contributing to calculations of
a Social Return Ratio and an SROI rate:
Time Period
0 1 2 3 4 5 6 7 8 9 10 Perp.
Present Value of the Benefits (NPV Bus. Cash Flow + NPV Social Benefits)
Present Value of the “Costs”* with IRR calculation provides:
Net Income of projected Social Benefits (Taxes from new Blended Value SROI
business performance + jobs, Government savings) =
attributed to investment -(minus) Social Costs or Total Business and
Social Benefits
* = Present Value of the “costs” in this case is the grant equity contributed to the organzation by government and foundation sources
9 Emerson, Jed, Jay Wachowicz, and Suzi Chun. “Social Return on Investment: Exploring Aspects of Value Creation in the Nonprofit Sector.” REDF
Box Set Vol. 2 – SROI Paper. 2000. Pg. 132-145.
19 measuring the “impact” in impact investing
SROI is much more than a single number. SROI builds on the social science data included in a typical cost
benefit analysis and should be considered as an entire analysis, rather than as a stand- alone figure. SROI
analysis is a way of reporting value creation over time. REDF prides itself in awarding funding in a “highly
investment-like way” based on these calculations.
Through the MJS Study’s January 2015 report “Economic Self-Sufficiency and Life Stability One Year After
Starting a Social Enterprise Job,” Mathematica introduces a slightly different way to calculate SROI worth
noting.
Using both collected information on real dollars (e.g., income, cash assistance) and “fixed effect models” (gen-
eralization of the difference-in-difference approach to convert, say, healthcare outputs to measurable dollars),
Mathematica conducted a Cost-Benefit analysis (CBA) to different groups of stakeholders. (Before this study,
REDF would measure benefits accrued based on REDF’s investment alone.) In this new study, while the
general calculation of SROI is similar, “the CBA explores the value of the average dollar spent by the social
enterprises funded by REDF from four perspectives: (1) society as a whole (the total benefits of the SE’s expen-
ditures), (2) SE workers (benefits to individuals served by the SE’s social mission), (3) the SE itself (as a busi-
ness venture), and (4) taxpayers not directly involved with SE (benefits to the community, excluding those
directly benefiting from the SE).”11 Only after each CBA is calculated does Mathematica bring together all costs
and benefits to calculate an overall SROI.
10 Rotz, Dana, Nan Maxwell and Adam Dunn. “Economic Self-Sufficiency and Life Stability One Year After Starting a Social Enterprise Job”.
Mathematica Policy Research Report. 2015. Pg xvii.
11 See 10, page 36.
20 measuring the “impact” in impact investing
GBF uses an Economic Rate of Return (ERR) calculation to understand the main economic benefits generated
by their portfolio companies and to distinguish the economic benefits accruing to low-income groups. They
calculate the ERR originally during the due diligence phase of the investment, and then recalculate it through-
out the investment life cycle.
GBF’s ERR calculation expresses a return on the total capital needed to operate a portfolio company (not just
GBF’s investment), using 10 years of economic projections. To calculate this, GBF includes the main econom-
ic benefits generated for various stakeholder groups, expressed as an annualized social and economic return
relative to the capital employed to produce that impact by applying an IRR calculation to the economic flows.
From these projections, GBF observes the total value generated by the company, as well as the portion of those
benefits that accrue to low-income stakeholders (suppliers and workers).
GBF revisits these projections on an annual basis, incorporating new information about client performance or
refining driving assumptions through survey work with the client. When possible, GBF cross-checks client-re-
ported data against the financial information they receive, supplemented by what they observe through work
with clients. GBF has also conducted supplier-level surveys with select clients to gather demographic information
about the suppliers and workers of their clients and to establish a baseline data set to track progress over time.
Based on similar principles as the SROI, Robin Hood Foundation uses Benefit-Cost ratio to capture a “best
estimate of the collective benefit to poor individuals that [their] grant creates per dollar cost to Robin Hood
– a direct analog to commercial return.”12 At its core, the BCR relies on translating the outcomes and typical
metrics of programs that can vary widely – from job training to pre-school to micro-lending – into monetized
values that measure poverty fighting. The BCR is calculated as follows:
(Poverty-Fighting Benefits of a Program / Costs to Robin Hood) x Robin Hood Factor
The numerator reflects a dollar estimate of the poverty-fighting benefits of the program to be funded, often
operationalized in terms of the private benefits that accrue to poor individuals over their lifetimes as a result;
the BCR leverages the expertise of Robin Hood’s program officers and social science literature to come up
with quantitative translations for benefits. The denominator represents the cost to Robin Hood of the grant.
The Robin Hood Factor is an estimate of the portion of the benefit that could reasonably be attributed to Robin
Hood’s funding. This takes into account the organization’s capacity to tap into alternative funding sources, and
the potential implications of Robin Hood not funding the organization. In effect, this considers the addition-
ality in Robin Hood’s role.
12 Weinstein, Michael M. and Cynthia Esposito Lamy. “Measuring Success: How Robin Hood Estimates the Impact of Grants”. February 2009. Pg 16.
21 measuring the “impact” in impact investing
BCRs help an organization determine which grants will yield the higher impact. Each BCR helps the Robin
Hood estimate the benefit of an investment “(measured in part by the projected boost in future earnings) that
each grant creates per dollar cost to Robin Hood.”13 This approach allows the Foundation to shift funds from
lower BCR programs to higher BCR programs, ensuring that the dollar can go as far as possible. As such, the
BCR is useful not only in estimating impact upfront, but in monitoring impact as well. In fact, the Robin Hood
computes the BCR on an ongoing basis and often doubles investments where the BCR is highest during the
re-investment / re-granting process.
For instance, as described in their 2009 report on “Measuring Success,” a BCR may help staff determine:
“whether to invest in a high school that graduates 50 former dropouts or to invest the same amount of
money in a training program that places an extra 75 workers in long-term jobs. The innovative methodolo-
gy compares the poverty-fighting value of any two grants, no matter how different in purpose. In effect, we
estimate benefit-cost ratios to compare the value of apples (graduating 50 more students from high school)
with the value of oranges (training an extra 75 home health aides).” 14
13 Weinstein, Michael M. and Cynthia Esposito Lamy. “Measuring Success: How Robin Hood Estimates the Impact of Grants”. Robin Hood
Foundation. February 2009. Pg 5.
14 See 13.
22 measuring the “impact” in impact investing
The expected return methodology can be perceived as – but this does not take into account how well the
inexact and constantly changing. Clearly, wide variabil- management team or entrepreneur may perform to
ity exists in ways to calculate expected return. Not only enhance the expected return.
can expected return take on various forms, but the cal- • Does not capture catalytic effect of investments. Some
culations themselves depend on what long-term societal investments can create change at scale by demonstrat-
savings are included in the calculations, for what period ing success of a disruptive model and inviting copycats
of time, and so on. For example, REDF’s SROI measure to enter the market, thereby driving change at scale.
has the tendency to focus attention upon cost savings to This is not typically factored into an expected return
society, but does not adequately incorporate many of the calculation, and would also be challenging to estimate
ways that employment improved individual peoples’ lives. in this method.
We define “impact” more specifically in this context than adjusted for what would have happened anyway, actions
we did for the overall purposes of the report as described of others & unintended consequences.” 16 Demonstrat-
in section 2.0. For our purposes in articulating the logic ing additionality seems to be key to EVPA’s definition of
model, impact refers to change occurring in communi- impact. EVPA refers to impact as “a technical and often
ties or systems. 15 This is not to be confused with the Euro- academic discussion,”17 and encourages impact investors
pean Venture Philanthropy Association definition, which to focus their efforts on outcomes.
describes the “impact” of a logic model as “outcomes
This difference in definition can lead to confusion for • Aligning incentives: The resulting map of outputs and
entrepreneurs who are building logic models with or for outcomes from the logic model can be used to identify
multiple funders; we use Kellogg’s definition and encour- targets upon which incentives systems can be designed.
age investors to clearly define what they refer to as Impacts This can be done to design incentive schemes for port-
when they are using the logic model framework, especial- folio managers based on social impact achieved (more
ly when communicating with external stakeholders. on incentives in section 5.1); it can also be used with the
investee management to set funding milestones based
on social impact objectives achieved as per targets
based on the logic model.
Application
• Reporting externally: The logic model framework is a
A logic model can be a simple and useful framework straightforward, simple way of illustrating an investee’s
to establish an investee’s path towards creating social path to creating social impact, even with audiences that
impact. The tool is often used in conjunction with other may not be familiar with sophisticated impact mea-
methodologies mentioned in this paper. Uses may surement methodologies. This can be used as a tool to
emphasize measurement of indicators at various parts communicate to external stakeholders about the theory
of the logic model. of change and progress.
18 Ebrahim, Alnoor and V. Kasturi Rangan. “The Limits of Nonprofit Impact: A Contingency Framework for Measuring Social Performance.”
Working Paper. Harvard Business School. May 2010.
25 measuring the “impact” in impact investing
Mapping the two dimensions against one another, • Niche results, where outputs may be very tangible but
Ebrahim and Rangan’s framework provides four quad- longer term impact is often beyond the control and role
rants. Assessing an investment using this framework of the specific organization (focused theory of change,
can be very useful in determining what level of the logic focused operational strategy); measurement should be
model would be most appropriate to measure. focused on inputs, activities, and outputs.
• Institutional results, where the organization has a • Integrated results, where the organization occupies
focused task (focused operational strategy) but impact several niches in the causal chain (complex operation-
is often achieved by networks rather than one organi- al strategy) and has a good level of control over both
zation alone (complex theory of change); measuring outputs and outcomes (focused theory of change);
outputs and intermediate outcomes are most appropri- appropriate levels of measurement may include aggre-
ate here. gate outputs, outcomes, and sometimes impacts.
• Ecosystem results, where the organization is aiming
for systemic changes (complex operational strategy) While the framework was developed with non-profit
and impacts are likely created through partnerships or impact in mind, it can also be very helpful in determining
a wide reaching breadth of services (complex theory of which level of the logic model an investor should focus on
change); outcomes and impacts are appropriate levels when measuring an investee’s impact.
of measurement.
Mapping out theory of change is core to how Acumen understands the depth dimension of their impact, which
Acumen describes as addressing the question of “how much and in what way has someone’s life changed?”
Acumen does this by employing their version of a logic model framework to map out an investee’s theory of
change. It is a core tool in thinking and communicating about their work’s depth of social impact.
Mapping the theory of change helps Acumen in identifying and predicting to what degree of the work by
investee impacts the lives of their target customers. In addition, Acumen highlights three other purposes of
using their theory of change framework:
• Identify assumptions: By digging deeper into the theory of change, Acumen is able to identify key assump-
tions required to get to the outcomes on the logic model. It uncovers the beliefs that the investment team
would have to hold to believe in the investee’s projected impact.
• Identify areas in need of further evidence review: Uncovering these key assumptions allows Acumen to iden-
tify areas that are in need of a further evidence review, and key data that they would like to collect over time
to verify that the theory of change holds. The Acumen team begins to answer this by looking at existing liter-
ature on the topic, and may combine it with primary research about the specific case in question if possible
to form an educated opinion.
• Consider impact risk: Acumen refers to impact risk as factors that could jeopardize the expected social impact
of an intervention, often found in the assumptions between “output” and “outcome”, and in the assump-
tions between “outcome” and “impact”. For each of their investments, the Acumen team outlines what they
think are the biggest impact risks, then come up with risk mitigation strategies to monitor and manage any
26 measuring the “impact” in impact investing
potential challenges. Understanding impact risk is seen as vital to measuring and ensuring long term
impact, and the up-front analysis allows the team to be better prepared.
The Acumen team has adopted their own version of a logic model that best suits their types of investments,
and also developed guidelines to identify common assumptions. This model is illustrated below:
Input: The primary Output: The product(s)/ Outcome: The result of Impact: The longer-term
product(s)/service(s) service(s) being consumed adopting the product/ effects on the target
offered by the organization/ at the household service expressed as customers’ household well-
initiative or customer level the monetary and non- being that can be attributed
(Measurable metrics) monetary well-being of to the good or service.
the target customers Impact will be (Measurable
(Measurable metrics) metrics when possible)
LGT Venture Philanthropy uses the logic model as depicted in the Kellogg Foundation Logic Model Devel-
opment Guide as a methodology. LGT combines this with the definition of Quality of Life of the UN Millen-
nium Ecosystem Assessment, and categorizes the outcomes of their investees using the five constituents of
Quality of Life: material well-being, physical well-being, social well-being, security, and freedom of choice.
Example: Drawing upon the European Venture Philanthropy Association, the table below illustrates the
logic model that LGT venture philanthropy developed for their investment in MFK, a ready to use food
(known as “RUF”) producer in Haiti. MFK produces fortified peanut based foods sold to institutional clients
who distribute them for free to malnourished children in Haiti. LGT Venture Philanthropy used the logic
model to understand MFK’s objectives, map their theory of change, and identify specific metrics for mea-
surement. The outcomes are overlaid with the five dimensions of quality of life, as consistent with LGTVP’s
overall methodology as mentioned above.
28 measuring the “impact” in impact investing
Figure 4.7 LGTVP’s Logic Model for MFK investment20 (adapted for length)
21 See 16.
30 measuring the “impact” in impact investing
metrics as discussed above, in scenarios where multiple indicators were identified; e.g., the prevalence of under-
funders and entrepreneurs are working to identify and weight children under five years of age and the propor-
select indicators to assess outcomes, we believe there tion of a population below the minimum level of dietary
is value in knowledge sharing among organizations to energy consumption.
learn from each other’s approaches and best practices
on leading indicators. For outcomes that are measured (2) Aside from applying a theory of change to assess the
across many organizations, there may be the potential to impact of an investee, the methodology may also have
adopt a common set of indicators. This can help lower application for impact investor funds themselves to use
the burden of measurement, and avoid reinventing at a portfolio level. In other words, this tool can be used
the wheel. Common indicators may be put forward by to answer the question: what is the impact investor’s
funders, or by a collective of organizations in the same own theory of change? This can help the impact invest-
industry. ing organization identify the hypotheses underlying its
investment thesis, and conduct deep dives to investigate
For example, Grameen Foundation’s Progress Out of such hypotheses to validate its own theory of change.
Poverty Index (PPI)22 is used by some organizations and As argued by Ebrahim and Rangan25, impacts are rarely
businesses with a mission to serve the poor. It can be achieved by individual organizations alone, but rather by
used to measure poverty-related outcomes; specifically, a number of interventions and actors that strive toward
poverty outreach, performance of an intervention among a common goal. An impact investor is uniquely situated
the poor, and track poverty levels over time. The PPI uses to see how the work of its portfolio organizations might
10 simple indicators that field workers can quickly collect link a series of outputs and outcomes that collective-
and verify. The answers to 10 questions about a house- ly lead to impact. Funders should consider articulating
hold’s characteristics and asset ownership are scored their own theory of change and assessing their own per-
to compute the likelihood that the household is living formance because it is at the level of the funder that sys-
below the poverty line. The PPI is country specific, and temic impact can be observed.
indicators are derived from the most recent national
household income or expenditure survey or the coun- For example, Root Capital’s theory of change is that
try-specific World Bank Living Standards Measurement investments in farming associations and private busi-
Survey.23 The 400-1000 indicators from the original sur- nesses can help build sustainable livelihoods for small-
veyed were narrowed down; statistics and expert judg- holder farmers26. As explained in Section 4.3, Root
ment constructed a 10 indicator scorecard that are linked Capital conducts studies to examine the assumptions
to the probabilities of rising above or falling below the underlying this hypothesis. Acumen also articulates a
poverty lines. We believe that there is potential in devel- theory of change for each of their investment sectors. For
oping similar types of indicator sets for measurement of instance, their theory of change for the health sector help
other common target social outcomes. to demonstrate how investment capital, sector expertise,
and industry linkages from Acumen can lead to desired
The United Nations Millennium Development Goals24 outputs, outcomes, and ultimately the target impact of
illustrate another example of a high level desired impact improved health and reduced disparities in healthcare
translated into specific indicators to demonstrate prog- delivery between rich and poor. This exercise can be
ress toward those goals. For instance, goal 1, to eradicate helpful in identifying or confirming key sub-sectors of
extreme poverty and hunger has three specific outcomes, investments, and can also be a useful communication tool
including to halve the proportion of people in the world for external funders who are interested in understanding
who suffer from hunger. For each outcome, a number of Acumen’s approach to creating impact in health.
22 “About the PPI: A Poverty Measurement Tool.” Progress out of Poverty Index, Grameen Foundation. progressoutofpoverty.org
23 “About LSMS.” World Bank Development Research, Living Standards Development Study. Econ.worldbank.org
24 United Nations Development Group. “Indicators for Monitoring the Millennium Development Goals.” United Nations. 2003.
25 See 4 above.
26 “A Roadmap to Impact.” Root Capital. 2012
31 measuring the “impact” in impact investing
Omidyar Network (ON) offers an example of an impact A theory of change mindset can also be useful for
investor that designed its impact measurement approach emerging impact investors to think through how they will
to reflect its own theory of change as an investor. ON’s drive impact through their investments and the evidence
impact measurement operates at two levels: firm level that they may wish to gather. For example, JP Morgan’s
impact (common for many impact investors), and sector Social Finance organization is currently working to
level impact. Sector level impact is a core part of ON’s build a knowledge base around social financial innova-
theory of change; in the organization’s 2012 “Priming tion. JP Morgan is measuring social impact of its early
the Pump”27 report, it argued that social impact at scale impact investments to enable them to gain expertise
requires not just investing in innovative firms, but also and a solid knowledge base to be able to later identify
fostering an ecosystem of multiple actors, including reg- the best opportunities for their clients. For example, JP
ulators and infrastructure providers. In addition to track- Morgan has launched a Social Finance unit which has
ing firm-level success, the ON team plans to use pub- focused on deploying proprietary capital to market-based
licly available information to assess progress of a sector solutions that can improve the livelihoods of low-income
in the regions in which they operate – for example; a and underserved populations globally. By measuring the
decrease in the number of unbanked individuals – and social impact of these investments, Social Finance is
also monitors intermediate milestones such as unlock- gaining a direct and detailed understanding of what type
ing of regulatory barriers. This allows the investment of investments are delivering the best results towards its
team to further understand the sector in which they are theory of change, and the learnings will inform future
operating and explore additional variables or nuances allocations for JP Morgan and their clients.
to add to their theory of change. “On Innovators and
Pinballs”28 published in the Stanford Social Innovation
Review offers more information on Omidyar’s approach
on measuring sector systems change and indirect impact
described here.
27 Bannick, Matt and Paula Goldman. “Priming the Pump: the Case for a Sector Based Approach to Impact Investing.” Omidyar Network,
September 2012.
28 Kubzansky, Mike and Paul Breloff. On Innovators and Pinballs. Stanford Social Innovation Review. September 2014.
32 measuring the “impact” in impact investing
Balanced
Scorecard 1. Learning 2. Business 3. Customer 4. Financial
Perspectives & Growth Processes
Logic
Inputs Activities Outputs Outcomes Impact
Model
Learning & Growth (L&G) Perspective objectives match metrics available as outputs or out-
The L&G perspective tends to describe how “people, comes in a logic model. It is important to note that social
technology, and the organizational climate”30 (including enterprises must think about two different sets of cus-
cultural attitudes) combine to support strategy and its tomers: beneficiaries (downward accountability), and
execution. In a logic model, this is similar to how one investors (upward accountability).
would think about inputs (people, technology, funding),
and how they combine to support activities of the orga- Financial Perspective
nization. Tracking outputs related to people, technology, The financial perspective includes financial data used to
and the organization can also help to inform effective- assess performance. As discussed in section 4.1, finan-
ness of learning and growth efforts. cial ratios can also be leveraged in the social sector to
estimate and later determine performance. (e.g., SROI).
Business Processes Perspective Using a logic model framework, financial performance
The business processes perspective “allows the man- is assessed starting from inputs, and then evaluated
agers to know how well their business is running, and against both outputs and estimated outcomes when
whether its products and services conform to customer possible.
requirements.”31 In a logic model, this corresponds to
thinking about how the activities of the organization are As a result of this analysis, we see how the logic model
supporting desired outputs. can be leveraged to build a scorecard analogous to the
one used by traditional business. In addition to the per-
Customer Perspective spectives included in the traditional balanced scorecard,
the logic model framework also offers an additional
A traditional scorecard uses the customer perspective
important link between outputs and impact. A potential
to measure customer-related objectives (e.g., customer
design is suggested in the “Recommendation” section.
acquisition, satisfaction, retention, profitability). These
30 Corporate Life Centre International Inc. “4. Learning & Growth Perspective” http://www.theclci.com/products_PMMS-BSC04.htm. 2015.
31 Balanced Scorecard Institute. “Balanced Scorecard Basics” http://balancedscorecard.org/Resources/About-the-Balanced-Scorecard. 2015.
34 measuring the “impact” in impact investing
Applications
As mentioned previously, mission alignment methods SF, New Profit), and 2) evaluate and monitor KPIs that
are useful in measuring execution against mission and track alignment with the investee’s mission and/or the
end goals. Specifically, mission alignment methods are investor’s investment thesis and mandate (e.g. New
used to 1) help ensure fit with the investor’s mission Profit, Acumen, Bridges Ventures).
during due diligence and investment selection (e.g. RSF
RSF Social Finance utilizes various forms and surveys to capture an organization’s estimated impact. While
one use of these metrics is to inform loan disbursement decisions, the primary use for the data is to monitor
RSF SF itself – the lender – and benchmark the progress of its portfolio against the mission.
Specifically, for a first time applicant, RSF Social Finance begins its assessments with a list of “Social Enter-
prise Mission Alignment Criteria” that both the organization requesting a loan and RSF SF complete about
the applicant. This is an example of a screen for ensuring mission alignment. The criteria include quanti-
tative questions around financial sustainability and public social/ environmental benefits of the program,
as well as more qualitative questions. RSF SF proceeds to evaluate the organization, by complementing the
form with other research, before presenting to a larger internal team who determines the amount and dura-
tion of the loan. The criteria are primarily used to compare a potential loan disbursement to their current
portfolio and ensure its mission alignment.
When deciding whether to extend or renew existing loans, RSF Social Finance requests that organizations
complete a “Portfolio Audit Survey” that they administer, as well as complete the B Impact Assessment. Again,
RSF Social Finance uses answers to these survey questions primarily as a way to monitor themselves, and a
way to benchmark their progress in supporting the organization.
(Note: The B Impact Assessment is a free, confidential tool powered by the nonprofit B Lab. The B Impact Assess-
ment enables organizations to assess how their company performs against dozens of best practices; compare its base-
line impact against other businesses; and develop a roadmap of improvements to deepen impact.)
35 measuring the “impact” in impact investing
New Profit utilizes a number of measurement methods – from scorecards to experimental evaluations – to
help its grantees scale, become financially sustainable, and maximize impact. Two of the most interesting
measurement methods New Profit utilizes to assess its impact are its externally-focused growth diagnostic
and its internal scorecard.
The external growth diagnostic allows portfolio managers “to consistently place organizations within the
growth stages”32 New Profit observes: startup, intermediary, or growth stage. These organizations, or invest-
ees, are classified into stages using 50+ dimensions (e.g., distribution model, costs, talent development).
This “stage” categorization helps portfolio managers to understand what growth targets and other KPIs to
expect. These are also the KPIs that feed into the internal scorecard, which is fairly straightforward and tracks
“activity on the deal” (KPIs) against New Profit’s goals.
It is important to note that, through an external assessment, New Profit measures its own contribution
effect. In a 4-year pilot study, New Profit has tracked its support activities on specific organizational dimen-
sions within a subset of grantee organizations and attempted to systematically associate those activities with
grantee self-reports of need, needs addressed, and growth/improvement. While the method has provided
valuable guidance for New Profit staff as well as evidence of its effects on grantees, using external evalua-
tors has made it resource intensive. The next phase of their exploration will be to test an internally managed
model that makes this level of monitoring sustainable, albeit at some loss of data independence.
Acumen’s focus is to invest in companies that increase access to basic goods and services for the poor.
Acumen’s impact philosophy considers that the ultimate measure of impact is how it changes lives at the
customer level, which they have defined as a function of three dimensions:
• Breadth: the number of lives reached and jobs created
• Depth: improvement in household wellbeing
• Poverty focus: who is being served
The Impact Template suggests similar common metrics to monitor impact for “Depth” and “Poverty Focus”
dimensions. Additional metrics may be added based on the investee’s theory of change (as described in
Section 4.2) and/or negotiations with the investee.
Bridges Ventures believes that understanding the risk associated with achieving social impact in an invest-
ment is just as important as the financial risk assessed pre-investment. Because of this, Bridges developed
what they call their “IMPACT Methodology,” comprised of the IMPACT Radar pre-investment, and the
IMPACT Scorecard post-investment. The IMPACT methodology is applied to any potential investment.
Pre-investment, Bridges Ventures utilizes an IMPACT Radar,34 which informs the due diligence of an oppor-
tunity. The IMPACT Radar lays out a high-medium-low risk/return analysis for each of the four investment
criteria that are core to its investment thesis: target outcomes, additionality, environmental, social, and gov-
ernance practices (ESG),
and alignment. The risk
and return of each invest- Figure 4.10 Bridges Ventures’ IMPACT Radar35
ment criterion is given
a high, medium, or low
score. Scoring is supported
by tools developed by the
firm that enable Bridges
Ventures to think about
each investment accord-
ing to similar frameworks.
The IMPACT Radar, along
with the tools that support
Bridges’ judgment in each
dimension of the Radar, are
illustrated in Figure 4.10:
Post-investment, the IMPACT Radar serves as a portfolio management tool through which Bridges can
“monitor the impact risk/return profile of each investment (and therefore of each fund) on an ongoing
basis.” Essentially, Bridges turns to using its IMPACT Scorecard to capture the key performance indicators
(KPIs) that relate to each dimension of the IMPACT Radar (e.g, there is a section for “ESG KPIs”). The KPIs
captured on a Scorecard are determined with each investee and reflect both outputs and outcomes of the
logic model. Designing the Scorecard in a way in which it links to the Radar (used during diligence) allows
the Bridges team to track performance of the investee in a thoughtful, consistent and organized way.
Bridges also uses the Scorecard to provide impact performance data (contribution to society) to investors,
alongside (separately sourced) financial performance (returns for investors).
Our Assessment impact or mask key issues with the social enterprise,
the mission alignment measurement fails to serve its
Advantages of Mission Alignment Measurement Methods purpose.
As mentioned in the Definition section, surveys and
screens are inexpensive, straightforward ways to monitor Additionally, depending on the customization of score-
mission alignment of an investor’s or a grantee’s impact. cards, they may not allow for direct comparisons across
While these forms of measurement can certainly take on different investments, as KPIs may vary across invest-
greater sophistication, at a basic level they can be utilized ments and interventions. This is in direct contrast to one
by an organization in any stage of development. RSF SF of the major advantages of the expected return measure-
is able to effectively utilize its social value criteria during ment methodology, which enables a common vocabulary
diligence and throughout the investment lifecycle. by translating impact into economic terms.
A suggested (and very oversimplified) scorecard set-up After the scorecard is set up for the impact investor at a
for an impact investor based on a logic model could be fund-level, a similar scorecard framework can be used for
the sample “Impact Investor Scorecard” in the top of each investee. Specifically, although the theory of change,
Figure 4.11. inputs, and activities would differ based on the invest-
ee, the investor’s theory of change can be used to help
The scorecard should first incorporate inputs and activ- identify common and specific output KPIs to be tracked
ities, and show how they translate to outputs and ulti- for each investee. This may be done through common
mately to outcomes. thematic categories of KPI’s that are fundamental to
the investor’s theory of change (for example, “breadth,”
As outcomes are more difficult to measure – especial- “depth,” and “poverty focus” as per the Acumen example
ly with any sort of frequency – “associated outcomes” above). Common thematic categories of KPIs can help
should be calculated based on output metrics and studies ensure all investments are in alignment with the impact
of causation that can be drawn from previous research or investor’s mission and theory of change.
other evidence.
Figure 4.11, which we referred to previously, also illus-
And finally, additionality should be estimated or trates the link between the two: the scorecard at the top
described – even if it is qualitative. This last step is import- shows the template for portfolio-level monitoring, which
ant, because the impact investor can then go through a is in effect an aggregate of KPIs tracked by an adapted
conscious exercise of determining to what degree the scorecard at an investment level shown below. While we
estimated additionality is in alignment with the organi- recognize that this illustration of metric aggregation is
zation’s theory of change. This is a key link missing in overly simplistic, we believe that pursuing an approach
traditional scorecards. Additionality is discussed further that uses investee information to form a portfolio view is
in section 5.2. worthwhile to enhance mission alignment.
Figure 4.11 Sample Investor-Level Scorecard Template and Hypothetical Portfolio-Level Scorecard Aggregation
}
<Organization (Investee) Name Here>
Low High
Inputs Outputs Associated outcomes
• Input #1 (e.g, # Thematic Category #1 (e.g., breadth) • Estimated outcome #1
<Organization (Investee) Name Here>
Activities
• KPI #1
• KPI #2
• KPI #3
Rating of Impact alignment
with Theory of Change
• KPI #2
•…
Low High
• Activity #1(activity • KPI #3 • Describe as best as
performed by impact •… possible the additionality
<Organization (Investee) Name Here>
Activities
• Activity #1(activity
• KPI #1
• KPI #2
• KPI #3
Additionality
• Describe as best as
• KPI #3 with Theory of Change
performed by impact
investor)
•…
(2) Secondly, while identifying the best KPIs can be a power of effective scorecards. While organizations have
complex exercise, organizations now have access to cat- access to suggested KPIs via networks like the IRIS and
alogs of generally-accepted performance metrics that GIIRS, it seems that the biggest gap right now – not so
leading impact investors use to measure social, envi- much for impact investors, but for investees – is utilizing
ronmental, and financial success, evaluate deals, and scorecards effectively and updating them often. Through
grow the credibility of the impact investing industry. impact investing forums, both investors and investees
The bigger challenge, we believe, is how to effective- should learn how to design and best populate scorecards;
ly manage this data and monitor the social impact and in fact, forums and other forms of educational events
mission alignment of investments over time. would be great times for established impact investors to
showcase scorecard templates. Once more impact inves-
As a result, we believe that there is a real opportunity to tors are on board they can offer technical assistance to
educate impact investors and their investees around the investees so these organizations can learn as well.
36 So, Ivy and Adam Jagelewski, “A. Social Impact Bond Guide for Service Providers.” MaRS Centre for Impact Investing. November 2013.
41 measuring the “impact” in impact investing
Pre/post: Comparison of intervention A group of marginalized women receive • This approach assumes that change
group before the intervention and after an intervention designed to increase was caused by the intervention, and is
it has concluded. The “pre” baseline their employability. The group is mea- unable to account for external factors
wave serves as a control group. sured for their employment rate before that may have also contributed to the
and after the intervention. change.
• While the extent of change to the
specific population is measured, it does
not reveal why conclusively.
Regression discontinuity design: Com- An intervention is targeted towards • This may incentivize service providers
parison with outcomes of those just be- students who scored below 50% on a to focus their efforts on those closest to
low or just above intervention eligibility diagnostic test. The evaluation frame- the threshold, even if those are not the
thresholds. This methodology is based work may be based on comparison clients that require the most effort.
on the premise that the difference of the post-intervention outcomes of • This methodology is only applicable in
between candidates who just miss and students who scored just above 50% interventions where the characteristics
just make a threshold are negligible, with those of the students who scored of those just above or just below the
and thus comparison of their outcomes just under 50% pre-intervention. threshold are negligible.
post-intervention reveals the impact of
the intervention.
Difference in differences comparison: Two prisons have historically shown • The treatment as usual group must
Comparison with a similar population – very similar recidivism trends. One be a good representation of what would
one that is not offered the new inter- prison is receiving the intervention, have happened to the treatment group,
vention, but which is receiving another while the other is receiving treatment as in the absence of the intervention.
“treatment as usual”. Both groups usual during that timeframe. The differ-
receive pre-and-post assessments, and ence in recidivism before and after the
the difference between those assess- intervention is calculated for the treated
ments is used to determine impact of population. The difference for the same
the new intervention. timeframe is also calculated for the
treatment as usual population. The
difference of these differences reveals
the impact of the intervention.
42 measuring the “impact” in impact investing
Both experimental and quasi-experimental methods have been used in the outcome measurement designs
of Social Impact Bonds.
New York State, randomized control trial: The first SIB to use an RCT in determining outcome payments was
the social impact partnership between the state of New York; Center for Employment Opportunities (CEO);
Bank of America Merrill Lynch; Robin Hood; Rockefeller Foundation; Chesapeake Research Associates; and
Social Finance US. The deal aims to expand comprehensive reentry employment services to 2,000 former-
ly incarcerated individuals in New York City and Rochester, New York. For the evaluation process, the New
York State Department of Corrections and Community Supervision identifies eligible individuals in NYC
and Rochester and randomly assigns them to the treatment and control groups. The performance-based pay-
ments will be based on three outcome metrics:
• Recidivism: The number of “bed days,” as captured in the NYS Department of Corrections and Community
Supervision administration data systems.
• Employment: Indication of positive earnings, as captured in NYS Department of Labor’s quarterly unem-
ployment insurance wage data.
• Engagement in transitional jobs: number of treatment group members who start a Centre of Employment
Opportunities transitional job, as captured by the CEO intervention data.
These metrics were selected by the project partners based on a set of criteria, including: the degree it rep-
resents meaningful improvement in the lives of individuals; alignment with the intervention’s theory of
change; relationship to public sector savings and other benefits; whether it is captured in existing state
administrative data systems; and degree that the outcome of the metric can be affected by the intervention,
as demonstrated by prior evaluations.
The Social Impact Bond was announced in December 2013 and the measurement and payment calculation
of Phase I is expected to take place in Year 4.
Essex County, historical baseline: The SIB will enable the Essex County Council to provide Multi-Systemic
Therapy (MST) to 380 young people at risk of entering care and their families over an 8 year period. MST is
an evidence-based program that delivers family therapy in the home through qualified therapists.
The primary metric on which outcomes are measured is the number of care placement days saved over 30
months post-referral as compared against a control review figure; i.e. the average number of days spent in care
by a comparable group of children over a 30-month period. This control review figure was established prior to
signing of contracts and is based on a historical case file review of 650 cases with data tracked over 30 months.
The outcomes contract was signed in November 2012, and the intervention is currently in progress. To date,
no outcomes data has been found for public release.37
37 “Social Impact Bond Knowledge Box.” UK Cabinet Office Center for Social Impact Bonds. data.gov.uk/sib_knowledge_box. Web. Accessed
December 2014.
44 measuring the “impact” in impact investing
Peterborough, UK, difference in difference: In September 2010 HMP Peterborough became the site of the
world’s first SIB, aimed at reducing re-offending rates among short-sentenced prisoners leaving Peterbor-
ough Prison. The aim is a reduction in court, police, and prison costs as a result of reduced re-offending, for
which reconviction events are a suitable intermediate proxy.
The success of this Social Impact Bond will be measured by reconviction events by all of the short sentenced
male prisoners from Peterborough prison — whether or not they engage with the service. Each cohort will be
compared by an independent evaluator to a similar group of prisoners across the UK from the Police Nation-
al Computer. The independent evaluator is responsible for developing the comparison group of prisoners
against which the SIB’s 3,000 short-sentence prisoners will be compared to assess whether the outcome
has been achieved and a payment is due to investors. The comparison group is developed using Propensity
Scoring Matching (PSM) methodology – this is where each Peterborough prison-leaver is matched to up to
10 prisoners released elsewhere in the country. This is done to remove, as far as possible, the influence of
external factors on reconviction levels.
Outcome payments will be made if there is a 10% reduction in the number of reconviction events over 12
months compared to a control group, or if the SIB’s three cohorts achieve an average reduction of 7.5%.
The first interim results, released August 2014, showed that there was an 8.4% reduction in re-offending
amongst the intervention group compared to the national average. The service reported that its success is
increasing over time as it gains experience and is able to learn from early challenges.38
38 See 37 above.
45 measuring the “impact” in impact investing
One of Root Capital’s main goals in measuring impact is to test their primary hypothesis that agricultural
businesses enable farmers to achieve improved livelihoods. They design studies of selected clients to eval-
uate whether and how their client agricultural businesses support farmer livelihoods, to verify that they are
truly reaching under-served businesses, and to inform their assumptions about what social and environmen-
tal practices truly create positive impacts.
In 2014, Root Capital released their first multi-site impact study - Improving Rural Livelihoods: A Study of
Four Guatemalan Coffee Cooperatives39. This study provides a detailed picture of the impact that Root Capital’s
client enterprises have on the livelihoods of smallholder framers and the environment, and seeks to answer
the question: Does Root Capital’s financing and training enable their clients to increase their impacts, and if
so, how and to what extent?
This study looked at four of Root Capital’s coffee cooperative clients in Guatemala. The research focused
on the cooperatives’ roles in promoting farmer livelihoods, and involved surveying 640 farmers. With each
cooperative, they recruited a comparison group of farmers living in the same communities to allow the
researchers to correlate differences (e.g., in income, access to services, and production practices) with ser-
vices provided by the cooperatives.
The “target outcomes” component of Bridge’s IMPACT Radar is concerned with the outcomes that the
investment is intended to generate, and the strength of the causal links in the investment’s logic model – i.e.
the extent to which causality has been evidenced from the venture’s progress thus far. In their framework,
Bridges scores an investment “low” on the target outcome risk scale if there is a minimal threat to the logic
model, from internal or external factors, such as if there is a scientific study (e.g. control trial or longitudinal
study, which is considered quasi-experimental for our purposes) that evidences causality. It is considered that
such a study demonstrates that the investment is generating impact. On the other hand, if the investment
only has secondary research that evidences causality in a different but comparable context, then it is regarded
as “high” target outcome impact risk on Bridges’ scale as it poses a high threat to the logic chain.
39 “Improving Rural Livelihoods: A Study of Four Guatemalan Coffee Cooperatives.” Root Capital. November 2014.
46 measuring the “impact” in impact investing
Acumen is a non-profit global venture fund that uses entrepreneurial approaches to solve the problems of poverty.
The aim is to help build financially sustainable organizations that deliver affordable goods and services that improve
the lives of the poor
LGTVP and Acumen are supporting an ongoing JPAL study to measure how much off-grid consumers in
Bihar are willing to pay for Husk Power electricity and whether their wellbeing changes as a result of having
access to energy. Husk Power is a portfolio company of both Acumen and LGTVP, and JPAL is a global
network of researchers who conduct randomized evaluations to test and improve the effectiveness of pro-
grams and policies aimed at reducing poverty. Studies such as this one helps increase the evidence level that
portfolio companies are driving impact, and also helps both the portfolio companies and the impact inves-
tors learn about and test the links of impact in the investee’s theory of change.
Similarly, LGT Venture Philanthropy funded an independent study of an education investment in India’s
rural and tribal communities using creative learning techniques that combine the traditional way of family
with education. This study revealed that while the creative learning techniques did achieve the desired
effect, they were most pronounced in the students that were already performing relatively well. This enabled
LGTVP to work with its investee to adjust the training of the educators to further improve the outcomes of
the students that they want to reach.
RCT’s are most appropriate when the causal variables 3. It may not be ethical or desirable to randomize people
and their effects can be clearly distinguished, and sample into a treatment and control group. Some have ethical
sizes are sufficiently large. Since RCTs are so resource concerns about denying the control group an interven-
intensive, there should be an amount of certainty that tion that could be helpful for the purposes of an evalu-
the evaluation will reveal statistically significant success ation exercise.
for it to be worth the investment. Additionally, because
an RCT cannot be adjusted midway through, course-cor- ….with Quasi-Experimental Methods
rections can be restricted due to the evaluation process.
While they are less costly than RCT, quasi-experimental
This limits intervention’s management ability to respond
methods still require significant effort for each under-
to changes in real time, which can lead to a very frustrat-
taking. And unlike an RCT, despite such effort there
ing work environment, as well as less than ideal treat-
can often be limits to their ability to rule out exogenous
ment for the beneficiaries involved.
factors, depending on the rigour of the counterfactual.
In general, RCTs are not suited for all interventions, for
Other thoughts
three distinct reasons:
Compared to most of the methodologies discussed on this
1. The evaluation must be able to be administered in an
paper, experimental and quasi-experimental methods
environment that can be somewhat controlled, which
are often costly and resource-intensive. As such, it is no
is not always possible. For instance, MDRC piloted an
surprise that while these methods are widely advocated
RCT for the Rikers Island SIB but found that the eval-
in the Social Impact Bond context, they do not appear to
uation would have been ineffective. Through this SIB,
be widely used by most other impact investors. Even in
MDRC is implementing a cognitive behavioral therapy
the cases where they are used to assess an investment’s
program for 16- to 18-year-olds detained at New York
impact risk, impact investors appear to seek out previous-
City’s Rikers Island with the goal of reducing the high
ly completed experimental or quasi-experimental studies
recidivism rate for this population. After the RCT pilot,
that demonstrates evidence for outcome, rather than
it was evident that the environment was too volatile
designing and executing a study themselves.
for the controls required by an RCT and the popula-
tion moved around too much to be able to measure the
While using these methods may be infeasible for all
impact of the intervention alone. MDRC had to settle
investees of an impact investment, we believe that these
for a pre-/post-comparison study instead.
methods hold great potential to test impact investors’
2. Few interventions are sufficient to drive outcomes on own theory of change. This is particularly relevant for
their own; these typically require comprehensive, wrap- impact investors that have a very focused investment
around solutions. Many social programs and enterpris- thesis driving towards specific change, as illustrated by
es work within an ecosystem, and isolating the impact the Root Capital example above.
of the specific investment can be very challenging and/
or inappropriate.
48 measuring the “impact” in impact investing
We believe that replicating these models in other sub- stock of existing available data as they begin designing the
sectors can be very valuable for impact investing and evi- experimental and quasi-experimental evaluations to lever-
dence-based decision making more broadly. age opportunities to lost cost and effort required.
(2) The cost and effort related to conducting experimen- Low-cost RCTs also suggest embedding random assign-
tal and quasi-experimental evaluations can be reduced by ment as part of usual program operations. The principle
following the principles of low-cost RCTs45, a recent inno- holds that since programs often do not have sufficient
vation in policy research that holds the potential to more funds to serve everyone who is eligible, program manag-
rapidly build evidence of “what works” to address social ers might as well use random assignment to determine
problems as proposed by the Coalition of Evidence-Based who will be offered program services, thus providing a
Policy. Low-cost RCT principles suggest using admin- base for a randomized evaluation. While this may not
istrative data that are collected for other purposes to be as applicable in impact investments where custom-
measure the key outcomes, rather than engaging in orig- ers are the beneficiaries – since beneficiary selection is
inal data collection. Particularly in more developed coun- largely driven by market forces in those instances – this
tries, administrative data may be available to measure out- may be relevant for Social Impact Bond interventions or
comes such as employment, earnings, student test scores, social enterprises that produce impact as a result of their
criminal arrests, receipt of government assistance, and operations (e.g. increasing opportunity by hiring mar-
health care expenditures. Evaluators can benefit by taking ginalized population as workforce).
5. Cross-Cutting Themes
In addition to the specific methodologies outlined, our Survey fatigue of beneficiaries
research also revealed two cross cutting themes: stake- In theory, impact measurement can help improve the
holder incentives and additionality. These themes will be investment’s effectiveness in creating social impact, and
explored more in depth in this section. thereby improve outcomes. However, this link is often
is lost on the beneficiaries who may be repeatedly bom-
barded with survey or interview requests. On BBC’s
5.1 Incentives “The Forum Program”46, Mike McCreless, Root Capi-
tal’s Director of Strategy & Impact, provided an anecdote
While impact measurement offers benefits for multiple that describes this issue:
stakeholders, there appears to be a need for greater align-
ment of incentives to devote the resources to measuring “There was a farmer that we talked to, we started
impact. Currently, there is little clarity or consensus on asking him the typical questions and he said, ‘You
whose responsibility it is to lead the impact measurement know, I answered a survey like this 6 months ago, 12
work. We believe that a number of factors contribute to months ago, 2 years ago and 3 years ago…people are
the lack of incentives to measure impact, including: always coming here to try to measure how poor are we.
All you’re going to do is take the data and take it to the
• Perception of low value in impact measurement by funding agencies and you’re going to get funded and
entrepreneur I’m going to be left with nothing so I’m not going to
• Survey fatigue of beneficiaries answer these questions and I’m not going to partici-
• Low fund investor appetite for robust measurement pate in this any longer, goodbye sir.’ He was tired of
• Limited incentive structures for delivery of social impact being measured.”
We dive into each of these below. As McCreless points out, it is important to remem-
ber that the communities we are working with are not
Perception of low value in impact measurement by objects of research or experiments to be measured. They
entrepreneurs are people with their own lives and experiences. There
are no direct incentives for them to take their time to
Social entrepreneurs have limited time and resources to complete surveys or interviews, though their informa-
allocate to extensive impact measurement. Some entre- tion is often critical to the data collection process.
preneurs may see impact measurement as outside of the
work of pursuing their business opportunity and per- Low fund investor appetite for robust measurement
ceive it as a top down requirement from investors that
does not provide a direct benefit. Given that the data col- Expectations of individual fund investors (i.e. Limited
lection for impact measurement often requires cooper- partners, or LPs) also influence the level of emphasis on
ation or self-reporting from the entrepreneurs, this per- impact measurement. One portfolio manager interview-
ception can often be an obstacle to implementing impact ee described his goal as “maximizing financial return
measurement work, particularly in the monitoring and while meeting the impact threshold.” Similarly, impact
evaluation phases. investing firms report that their funders care that firms
have “metrics themselves, [but do not care] what the
metrics are.”47 Additionally, firms report that funders
wrongfully associate general metrics or financial returns
for social impact. In either case, there is a perception that
46 “Mike McCreless on the mango farmer who got fed up with completing surveys.” The Forum, BBC World Service. October 2013.
47 Excerpt from our interviews with impact investors. See Appendix A
51 measuring the “impact” in impact investing
the investors in the fund care less about the actual out- financially rewarded for delivery of social outcomes; their
comes from the investments than the practices and dil- performance is linked to their ability to deliver financial
igence around impact, potentially due to the amount of return.
effort required to fully understand the former.
As a result, in already tight budgets, impact measure-
We believe that there are a few underlying causes for ment is de-prioritized. Unless there is funding for mea-
low investor appetite when it comes to social impact surement, it can be a struggle to execute on the desire
measurement: to measure impact rigorously – especially because it
1. Fund investors believe that due diligence in the invest- entails not only coming up with the funding to do it, but
ment manager is enough. The investors or LPs in this also building the capacity to become more results-fo-
context were described as investors who seem satis- cused (e.g., infrastructure, technology, resources). One
fied that the fund managers demonstrate that they care of our portfolio manager interviewees said it best: “Until
about impact, by having methodologies and metrics someone gets paid for impact, measurement will lag. I
that embedded in their work. Plus, traditional investor don’t get paid to maximize impact return. I get paid to
clients – especially those that are less educated in finan- check the impact box.”
cial markets – perform diligence in finding fund man-
agers and rarely question competency after the initial
selection. Recommendations
2. The private sector doesn’t operate like this. The private Many observers believe that impact investing is at an
sector does not have to justify impact to investors; important juncture as it tries to “make the move from
impact is measured with financial returns alone, and philanthropic thought experiment to powerful instru-
fund investors rarely assess a whole slew of other ment for global change.”48 Studies show that millenni-
outputs and outcomes if all holds steady. As a result, it als’ values, experiences, and preferences are poised to
is likely that some clients or investors do not even know accelerate impact investing, directing billions of dollars
to ask for social impact measures. toward social benefit.49
3. It is perceived that costly measurement activities are not
worth investor funds. Some LPs question whether invest- This potential influx of new investors, however, must be
ing to develop impact systems is the most effective met by an impact investing industry with a strong infra-
use of their funds, when these monies could instead structure. The ability to accurately assess social impact is
be used to finance further expansion or growth. Many a cornerstone of this necessary infrastructure. Further,
investors hope that their investees would take it upon without increased rigor in impact measurement and a
themselves to perform better impact measurement push to maximize impact, impact investing runs the risk
in improving their strategies and operations (Reed et of becoming a term being used merely as a marketing
al, 2014). There is also a view held by some investors tool. Finally, impact measurement plays a critical role in
that impact investing funds are penalized because they building the capital markets that reward performance.
have to incur costs related to impact measurement that For these reasons and others, we must solve the problem
traditional funds do not. of limited incentives around impact measurement in the
market today.
Limited individual performance incentives for delivery of
Take a survey respondent-centric approach
social impact
A respondent-centric approach can mitigate some of the
Lastly, there appear to be limited incentive structures at
challenges related to both entrepreneurs’ perception of
the fund manager level to reward the delivery of social
impact measurement and beneficiaries’ survey fatigue.
impact. Very few – if any – portfolio managers are
A number of investors highlighted the importance of
48 Emerson, Jed and Lindsay Norcott. “Millennials Will Bring Impact Investing Mainstream.” Stanford Social Innovation Review. April 2014.
49 Dhar, Vilas and Julia Fetherston. “Impact Investing Needs Millennials.” Harvard Business Review. October 2014.
52 measuring the “impact” in impact investing
thinking in the entrepreneurs’ shoes when designing financial compensation to provide a clear incentive for
impact measurement processes. For example, what data the GP to manage the portfolio toward high social as well
would they find valuable and meaningful? How is the as financial performance. An Impact Audit Committee
company and its products changing lives? This approach determines an Impact Score by reviewing the social per-
can provide customer insight for the company while also formance of investments and the GP’s actions in sup-
providing useful impact data. porting intended social outcomes. The GP’s annual
bonus, derived from a 2% management fee, is paid to
We have borrowed the term “respondent-centric the GP in direct proportion to the percentage of the total
approach” from Root Capital’s “client centric approach.”50 Impact Score achieved. The 20% carry is comprised of
Root Capital’s guiding principle is to generate data that two components: 90% of the carry is tied to financial
helps Root Capital understand its impact on small-scale performance, and the remaining 10% is tied to to the
farmers while also creating value for the farmers and Impact Score.51
enterprises. This often translates to generating the data
that the investor needs by working with clients and their Embed the impact measurement role
procedures to generate data that they need. As such, In some impact investing funds, investment managers
Root Capital positions itself as a value-added partner who tend to be drawn from the mainstream finance sector,
observes and measures impact to help farmers and enter- with impact assessment being left to a dedicated pro-
prises increase their value, rather than as impartial out- fessional who may sit alongside, or more often outside
siders measuring impact. Similarly, a respondent-centric the core team managing the investments.52 This practice
approach can improve the quality of the impact data by structurally positions impact measurement as an after-
focusing on issues of true importance to the communi- thought. Instead of concentrating the work of impact
ty, increasing the commitment level of participants, and measurement to a dedicated resource outside of the
creating an environment for honest and representative core investment process, impact investors should con-
responses. sider training their portfolio and investment managers
in impact measurement, and/or including the activity as
Design innovative incentive structures part of their core work.
Some impact investing funds are exploring innovative
incentive structures such as a “social impact carry,” For example, at LGT Venture Philanthropy, investment
where portfolio managers (and/or relationship manag- managers are responsible for impact measurement,
ers) are rewarded partly based on results related to mea- including building the theory of change, conducting site
sured social impact from investments under their man- visits and working with the ventures to collect impact
agement. For instance, they may be rewarded based on data. The Head of Impact Management works with each
the degree to which their investments meet or exceed the investment managers to review investment memos
predetermined impact targets. Similarly, the financial and other documents to provide feedback, and offers a
carry may be restructured to incorporate an element that sounding board and challenge function for investment
is contingent on social impact. managers throughout the process. The Head of Impact
Management also consolidates the impact figures to
For example, Core Innovation Capital created a direct report the data to stakeholders and looks to continuously
link between an Impact Score and General Partner (GP) improve data quality for the organization.
53 Brest, Paul. “The G8 Task Force Report: Making Impact or Making Believe?” Stanford Social Innovation Review. October 2014.
Best, Paul and Kelly Born. “Unpacking the Impact in Impact Investing.” Stanford Social Innovation Review. August 2013.
54 measuring the “impact” in impact investing
challenging to implement due to the significant disin- investors such as Bridges are already doing this; the spot-
centive for impact investors who currently offer mainly light below outlines their approach and framework. We
“red” products to self-identify their products as such. believe that a similar approach should be applied to other
Without a large group of investors willing to self-iden- impact investors to bring the subject of additionality into
tify as belonging an undesirable category, this system the forefront. We also recommend elevating the conver-
may find itself challenged to gather momentum. While sation on additionality and encourage it to be on the fore-
there may be an opportunity for an organization like the front of any conversation regarding measuring impact in
Global Impact Investing Network to take a leadership impact investing; this will help LP’s and other funders
role in introducing it into the ecosystem, our hypothesis in the ecosystem understand the concept and push for
is that the market institutions are not currently at a stage it to be included in the impact reports that they receive.
where this can be strongly enforced across the industry. Ultimately, we hope that this emphasis will push for a
greater emphasis on additionality, which in turn will lead
Instead, our recommendation is to encourage impact to a greater impact to addressing our society’s pressing
investors to try to measure the additionality in their social and environmental issues.
impact, and to include it in their impact reports. Some
Bridges’ additionality scoring guide considers both investor-level and enterprise-level additionality.
In their framework, an investment scores low on investor-level additionality if the business is already well-es-
tablished with other competing investors. In co-investment situations, investor-level additionality is analyzed
by the extent to which Bridges leads the development of the investment, and therefore the leverage of addi-
tional capital. An investment is considered medium in investor-level additionality if Bridges is the sole or lead
investor in an opportunity overlooked by mainstream investors. For example, Bridges’ investor-level addition-
ality in their Underserved Markets theme lies in directing capital to businesses that demonstrate strong value
to some of the most deprived communities in the UK. Another example is Bridges’ Social Sector Funds, which
provide flexible capital to business models that cannot attract commercial capital due to their structure or
target market. In addition to providing capital, Bridges’ non-monetary support that can drive increased impact
is also considered as additionality. The highest level of investor-level additionality is when Bridges is incubat-
ing the business in-house. For example, Bridges identified a gap in low-cost gyms in inner city areas, despite
the potential for exercise to address the rapidly rising levels of obesity and other chronic diseases. As a result,
the Gym – a chain of low-cost gyms in UK’s deprived areas – was incubated within Bridges. This involved con-
ducting due diligence research, writing up a business plan, selecting a management team, and executing the
launch. The Gym was founded in 2007, with Bridges holding majority ownership.
For the enterprise-level additionality aspect, Bridges assesses whether the social outcomes generated by
the underlying investment will create a positive net benefit for society, rather than displacing comparable
benefits in the current environment. Enterprise-level additionality is considered low if there is a likely dis-
placement of comparable societal benefits; for instance, if the investee is simply stealing market share with
no impact value-add. In contrast, enterprise-level additionality is considered high if displacement of com-
parable societal benefits is very unlikely due to increased quantity or quality by the investee in addressing a
current market failure.
55 measuring the “impact” in impact investing
Note: Investee maturity should be determined by the impact investor based on the investee’s size,
reach, budget, or years in existence
56 measuring the “impact” in impact investing
The best practices integrated model In post-investment, the investor works with the invest-
ee to gather data on the KPIs, and analyzes them to
Our best practices model is most appropriate for a monitor the social impact performance of the investee.
mature impact investor that is working with a sophisti- This information may be used to extract lessons learned,
cated investee. In this model, we envision that the inves- make course-corrections, and/or inform the investor’s
tor uses a number of tools to screen investees candidates broader strategy.
and conduct due diligence of potential investments.
In the evaluation stage, measuring social impact may
To begin the due diligence process, an expected return entail a quasi-experimental method evaluation; if required
calculation (e.g. SROI) is used to compare the impact – or if the impact investor is interested – we recommend
of potential investments. However, this is only one using some of the least resource-intensive quasi-exper-
of the several sources of input in the impact estimate. imental studies (e.g., pre-/post-test or historical baseline
The investor also works with the entrepreneur to map study). The findings from this are used to test the links of
out the logic model of their theory of change to under- impact in the investee’s theory of change, and build confi-
stand how the investment will convert theory to action. dence for similar business model for future investments.
This tool is also used to identify the causal links under- Note that RCTs are not recommended, even in the best
lying the investment’s path to social impact, and allows practices model. This is because, based on our research
the investor to identify hypotheses to test and assesses and analysis, we conclude that RCTs are often too academ-
the strengths of these linkages. This assessment may ic and resource-intensive to provide the value that impact
draw upon existing experimental or quasi-experimental investors look for in evaluating outcomes. Quasi-experi-
studies from a “what works” database that demonstrate mental methods are a superior use of funds in this case.
evidence of the causal links in impact.
Above this investee-specific level, we also encourage
Also in the due diligence and selection processes the inves- impact investors to map their own theory of change.
tor considers the enterprise additionality – whether the This exercise allows the investor to articulate and under-
social outcomes generated by the underlying investment stand how their investments translate into intended
will create a positive net benefit for society, rather than dis- impact. This process can also be used to identify hypoth-
placing comparable benefits in the current environment eses underlying the investment thesis, and the investor
– and investor-level additionality; i.e., the ease of capital may choose to conduct deep dives to investigate whether
for the investee, and any non-monetary benefits that the hypotheses confirm (or adjust) its own theory of change.
investor can offer to boost the investee’s social impact.
Finally, in this best practice scenario we envision that the
In the pre-approval stage, the investor works with invest- impact measurement efforts are well integrated into the
ee to determine KPIs to track on the scorecard used to portfolio team as part of the investor’s core work. Addi-
monitor the investment. These KPIs can be drawn from tionally, we envision that portfolio managers are reward-
indicators suggested by the logic model from the due dil- ed based on results related to social impact generated by
igence stage; the Contingency Framework for Measuring investments under their management, potentially in an
Social Performance is used as a basis for discussion on arrangement such as a social impact carry or bonus.
the level of KPIs used (e.g. outputs vs. outcomes). Strate-
gy maps or the investor’s scorecard template may also be
used as a starting point for identifying appropriate KPIs. The simpler adaptation
Regardless, they are negotiated between the parties and Recognizing that not all investors are ready to take on
selected at this stage. These KPIs should be chosen from all of the above, our framework proposes a simpler
a respondent-centric perspective so that the data gath- version for those that are just starting out. As a first step,
ered is useful to both the investor and the actor that is we encourage investors to work with entrepreneurs to
collecting and/or reporting the data (investee, beneficia- develop a logic model to map out their venture’s theory
ry / customer). Note that scorecards are recommended of change, so that the investor can understand and eval-
here over social value criteria; this is because we believe uate its path to impact. In the pre-approval stage, we
scorecards are a more robust form of performance mea- suggest adopting social value criteria to rate investments,
surement and tracking over time. and to monitor the investee’s progress post-investment.
57 measuring the “impact” in impact investing
7. Conclusion
The field of impact investing is attracting increasing We believe that informal, inconsistent, and weak impact
interest from investors, creating a greater number of measurement methods could be a real constraint to the
impact investing organizations, and fueling an inflow of growth of the impact investing sector and its prospects to
capital to the sector. The sector growth to date – as well create real social change. We believe that impact invest-
as its projected scale in the next 5 to 10 years – has led ing holds tremendous potential in tackling some of
investment stakeholders to pursue impact measurement our world’s most pressing challenges; however, we also
to understand both financial return and social impact. believe that the term “impact investing” runs the risk of
being diluted and used as a marketing tool if a certain
The aim of our study was to deepen understanding of level of rigor in impact measurement is not established
the specific practices and methodologies that established in the industry. To that end, we hope that this paper has
impact investors and other funders are using to measure contributed to the dialogue and progress development
the social impact generated by their investments, and to of impact measurement in the emerging impact invest-
analyze the conditions under which each measurement ment field.
method is most applicable.
58 measuring the “impact” in impact investing
Special thanks to our faculty advisor, Alnoor Ebrahim, for his continued support and valuable feedback in the development of this
report. We would also like to acknowledge the HBS Social Enterprise Initiative for making the production of this report possible.