Impact Report API

Updated February 2019

ImpactMatters offers access to a dataset of cost-effectiveness estimates for programs run by nonprofits. We work with nonprofits to self-report this data, guiding them through specifying an outcome, reporting outcomes data, accounting for attribution and comparing to costs. Here is a selection of impact reports:

This document describes an API that serves impact report data. We encourage partners to use this data to inform donors about the impact of their gifts and steer them toward nonprofits that best advance the donor’s mission. If preferable, we offer monthly data exports in spreadsheet form.

For questions or to get access to the API, contact Elijah at egoldberg@impactm.org.

Data use

Impact data has the potential to be misinterpreted. To minimize these effects, we require partners to abide by a set of guidelines when presenting impact report data. These guidelines are described in each data use agreement between ImpactMatters and a partner.

Overview

The current version of the API lives at https://report.impactm.org/api/v1/. We recommend presenting impact reports in a similar fashion to Charity Navigator. Here is an example.

Authentication is set using api_token.

Versions

Version Date Changes
Version 1 2/1/2019 Initial deployment

We intend to release a second version of the dataset by mid-2019 that includes a richer set of fields for reports. We welcome suggestions for data to include.

Endpoints

Endpoint Description
/reports/list Returns list of all published impact reports.

Field Reference

Field

Description

id

Unique identifier for the report

  • vddqsaawyy
  • 1r0pnr2uxi

nonprofit_name

Nonprofit that operates the program analyzed in the report. The name is specified by the nonprofit and so may not match the legal name or a registered Doing Business As name. Use the EIN for unique identification of the nonprofit.

  • Living Goods
  • BRAC USA

ein

Employment identification number for the nonprofit. The API only returns U.S. 501(c)3-registered organizations.

  • 20-5010527
  • 20-8456741

program_name

Name referring to the activities analyzed in the report. Where a nonprofit only operates one program and the program does not have its own name, the name of the nonprofit is used.

  • Community Health Workers Program
  • Empowerment and Livelihoods for Adolescents

program_mission

Describes the objective of the program in terms of its outcome. There should be a straight connection between the mission set forth in this field and the results reported in the impact statement field. Maximimum 250 characters. Example:

  • To empower community health workers with the digital tools and supervision to address basic health challenges.
  • To improve the lives of adolescent girls along multiple dimensions, including earnings.

program_activities

Concretely describes what the program does. The statement includes activities, such as "conduct business training", and avoids aspirational or general language. Maximum 250 characters. Example:

  • Living Goods trains women from local communities in Uganda to diagnose and treat malnutrition, infection and many common and life-threatening illnesses.
  • In girls-only safe spaces, or Empowerment and Livelihoods for Adolescents (ELA) Clubs, girls receive trainings in life skills, confidence building, financial literacy, and specific livelihoods, and microloans for older girls to start businesses.

impact_statement

Estimate of outcomes generated by the program, compared to cost. Estimate is net of counterfactual impact. Estimates are always in the form of a single sentence.

Every statement includes the outcome measured. Outcomes are concrete and tangible so that an outsider would understand the statement without additional context. Examples:

  • Year of healthy life
  • Income

Every statement also includes the type of participant benefiting. These describe general types, not the specific population served. Examples:

  • Child under 5
  • Low-income adult

Of course, central to impact statements is the cost-effectiveness estimate. These estimates are presented in one of two forms. For outcomes that are discrete, such as a life saved, impact is presented as the cost to achieve a single outcome. For outcomes that are continuous, such as income, impact is presented as the average cost per participant and the average impact per participant. Examples of complete statements:

  • Discrete outcome: $5,000 saves the life a child under the age of five.
  • Continuous outcome: $40 increases the future earnings of an adolescent girl by $200.

confidence_in_estimate

Available only for audited results. The field provides the rating out of five for the confidence in the estimate presented in the impact report.

  • ImpactMatters assigned a rating of 5/5 for the audit team's confidence in its impact estimate for the Community Health Workers program.

methodology

A mad-lib sentence describing the type of data. If the report was impact audited, it states that. If not and the report is based on a study, it states the study type. Otherwise, it states whether the data was pre-post or post only and the method used to attribute results.

  • Living Goods underwent a rigorous audit of its impact by ImpactMatters.
  • BRAC USA and third­-party evaluators conducted a randomized controlled trial of the ELA program to estimate impact.

data_source

Specifies how impact data was collected, written by the nonprofit with some guidelines. Maximum 300 characters.

  • Randomized controlled trial (R.C.T.) conducted by third-party evaluators Abdul Latif Jameel Poverty Action Lab (J-PAL) and Innovations for Poverty Action (I.P.A.) on the Living Goods program in Uganda.
  • BRAC USA's survey data collected directly from program participants in 100 treatment communities and non­-participants in 50 control communities.

attribution_method

Specifies how the counterfactual was estimated to attribute impact, written by the nonprofit with some guidelines. Maximum 300 characters.

  • ImpactMatters used J-PAL/I.P.A.'s estimate of the counterfactual mortality rate in areas that were not exposed to Living Goods' intervention for the relevant period.
  • BRAC USA estimated counterfactual success by tracking the outcomes of a control group whose members were randomly assigned not to receive the intervention.

details

Lists who created the report, with what data and how; and what dates of program implementation are covered by the report.

  • This report was independently audited by ImpactMatters. This report covers January 2014 to December 2016.
  • This report was created by BRAC USA with self­-reported data using ImpactMatters' impact reporting platform. This report covers May 2008 to June 2012.

is_audited

ImpactMatters conducts impact audits of some estimates and asssigns a confidence rating. If ImpactMatters has conducted an audit, this will be marked true. Otherwise, it will be left null.