City Resources/Interest Alignment. CANOPY is a tool useful for specifically large-scale resource allocation problems related to maximizing take-up of needs-appropriate services by focal populations. Specifically, the key components of appropriate applications are:

1.         Decisionmaker/planner. A single large decisionmaker, or a group of decisionmakers who are acting together towards common goals, who oversee provision of a large proportion of the type of service in question in their jurisdiction.

2.         Complexity of take-up. CANOPY is most meaningfully applied to situations where service take-up depends on a large number of household, agency, and neighborhood factors, and where service can be provided at many different locations and/or different configurations.

3.         Historical information about service take-up. CANOPY requires an ability to predict how take-up of services would shift in response to potential changes. This prediction must be developed by analyzing patterns of historical take-up decisions, ideally from the same setting (i.e. same city, and same populations) as involved in the eventual application.

As an example, the proposed proof of concept project involves partnering with city planners in Chicago, who are looking to maximize take-up of Head Start services by households with the greatest levels of need. The challenge is to identify how to divide limited funding among a diversity of agencies throughout Chicago, where households with eligible children may opt in or out, and have the potential to enroll with a range of different agencies that might meet their needs. Other examples might include decisions around city-wide provision of after-school or youth violence prevention services, including both funding of new slots, or expansion of new centers.

Examples of cases where CANOPY is not helpful, or not necessary, include:

●            The choice of which among a small number of youth programs should receive awards from a small foundation, where questions about impact of program funding on the full ecosystem of service are arguably less important than quality of services and track record of the applicant organizations

●            The choice of where to locate a single or small number of pilot sites by a local government for a new type of service, since (1) there is no track record of prior take-up that can be studied to imagine how this service might be taken up in the future; and (2) the small scale of implementation can be best informed by direct investigation of each of the fewer choices that might be made about type and location of services.

Data Needs/Prerequisites. A specific set of data types and structures is necessary for CANOPY to perform analysis of service take-up patterns that can be useful in recommending improved configurations for service provision. The table below describes the canonical data requirements for the broad class of potential applications, as well as data sources for a prospective proof of concept for planning Head Start services.*

*Note that Chapin Hall has data sharing agreements and/or formal partnerships to draw on to obtain access to these data sources listed for a project focusing on the city of Chicago.

Data Requirement

Data Source for Proposed Proof of Concept

Individual-level data on the full population to be served

Supplemental Nutrition Assistance Program (SNAP) and Internal Revenue Service (IRS) data on all income-eligible children ages 0-2 (Early Head Start) and (3-5) Head Start) in Chicago

Human service providers which can be chosen, typically at the facility-by-location level

Provider data from the Chicago Department of Family and Support Services (DFSS), with information on school- vs community basis, size, blend of funding, full- or part-day status

Historic choices made by those taking up services

Head Start enrollment data, by child and provider, from DFSS

Additional environmental information related to choice

Data on public transportation, neighborhood safety, proximity of providers to commercial areas, etc