> For the complete documentation index, see [llms.txt](https://ambitious-impact.gitbook.io/ambitious-impact-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ambitious-impact.gitbook.io/ambitious-impact-docs/explainers-and-house-views/our-charity-taxonomy/guidance-to-adjust-research-for-system-level-interventions.md). # Guidance to adjust research for system-level interventions Contributors: Morgan Fairless ## Our evaluation approach should vary according to the type of mechanism being evaluated * Ideas can be characterized by the [primary mechanism](/ambitious-impact-docs/explainers-and-house-views/our-charity-taxonomy.md) in their theory of change. * Researchers should be mindful of this position when deciding what to front-load and how to approach the research. * This guidance focuses on how to approach ideas that work on systems-level change. ## System-level change A system is a network of actors, characterized by their incentives, behaviors, and relations. A system does not need to be massive or global in scale (i.e., there are family systems, organizational systems etc.), but we normally think of system-level interventions as those tackling fairly large and complex systems. *For instance, working on training healthcare workers in the facilities they work in is probably best evaluated through the lens of a provider-engagement intervention, but working on changing the curriculum requirements for continuous professional development for healthcare workers in a country may fit systems-level change better.*
MechanismDefinitionHow to tell if it belongsExamplesCover image
Government policy, regulation and governance

Changing government's laws, enforcement, decision-making/norms or public spending


Is the primary lever a government or public institution changing the rules, their enforcement, or how public resources are used?Animal Policy International, Concentric Policiesaccount_balance_90dp_801C31_FILL0_wght400_GRAD0_opsz48.png
Market ShapingChanging market incentives so beneficial goods and services become more available, attractive, or adoptedIs the primary lever changing market conditions that many buyers, suppliers, producers, or consumers respond to?ARMoRprice_change_90dp_801C31_FILL0_wght400_GRAD0_opsz48.png
Information and evidenceCorrecting informational asymmetries, generating or communicating evidence Is the primary lever giving decision-makers better information, rather than directly changing incentives, rules, or implementation capacity?Lead Research Alliance, Happier Lives Institutefact_check_90dp_801C31_FILL0_wght400_GRAD0_opsz48.png
Tech development and adoption Developing, adapting, commercializing, or scaling a beneficial technology or product to increase availability or adoptionDoes the org do good by addressing gaps where technology/products don’t exist, aren’t good enough, are too expensive, or aren’t being adopted?Developing new fish stunners, Scaling in-ovo sexing adoptiondeveloper_board_90dp_801C31_FILL0_wght400_GRAD0_opsz48.png
The literature on systems-level thinking tends to be (in my opinion) quite difficult to parse and rife with unfalsifiable claims. We can make some broad generalizations, and draw from different theories, but try to avoid falling into the trap of taking any given theory of systems as standalone, capital T, Truth. {% hint style="info" %} Complex systems tend to show some recurring dynamics that are worth keeping in mind. The UK Government's magenta book has a useful summary in [table 1](https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/879437/Magenta_Book_supplementary_guide._Handling_Complexity_in_policy_evaluation.pdf) ( pages 12-16). {% endhint %} ## General advice ### Key areas of focus / things to frontload
Map the system and its dynamics {% embed url="" %} Other guidance: [here](https://www.thinknpc.org/resource-hub/systems-practice-toolkit/systems-mapping/) and [here](https://medium.com/@adam.d.groves/from-service-design-to-systems-change-72fa62b1714c)
Identify leverage points A leverage point is a concrete "place" in the system where one can intervene to get a larger overall effect in how the system as a whole behaves. The [theory of change](/ambitious-impact-docs/ambitious-impact-eval/theories-of-change.md) for a systems change idea we are evaluating should explicitly target a leverage point. As such, we should probably identify the target early on. Examples of leverage points: * The rules of the system (e.g., trying to change laws or regulations) * Mindsets or goals (e.g., expanding the moral circle to consider certain animals as morally relevant) * Feedback loops (e.g., improving the availability or quality of information) * Coordination (e.g., convening actors that do not usually coordinate) * Resources (e.g., allocation of finances towards a ministry)
Talk to people It is very difficult to understand a system and its components without speaking to people involved in it. Even if they are informal conversations, its important to try to cover as much of the system as possible. For instance, if you are evaluating a market, you probably want to buyers, sellers, and intermediaries.
### Reporting standards * The theory of change section should include the following sub-sections: * [Stakeholder mapping](/ambitious-impact-docs/ambitious-impact-eval/stakeholder-mapping.md) is mandatory for system-level ideas. You can include it under the system map. * Identified leverage points and a rationale for why we think they are key to target. * [Cost-effectiveness analysis](/ambitious-impact-docs/ambitious-impact-eval/cost-effectiveness-modelling.md) should be carried out to our back-of-the-envelope standard. * [Neglectedness](/ambitious-impact-docs/ambitious-impact-eval/assessing-neglectedness.md) section should speak to attribution[^1], and explain under which conditions we expect the efforts to be additional. *** ## Pointers to evaluate different mechanisms
'Government policy, regulation and governance' ideas * **Core cruxes to investigate** * Aim to have a good understanding of the core "streams" involved in the target change * What are the politics of the issue? * What is currently being discussed in the "policy stream" (i.e., what solutions, ideas, and topics are discussed as part of debates on this topic; what is the option set)? * How is the problem you are trying to solve, and/or the solutions that you are trying to focus on, framed? * What could a window of opportunity in this space look like? What would the organization try to leverage? * What coalitions are usually involved in this issue? * How does the issue, its framing, or its politics, change among different countries? What stays the same? * Be critical of having a bias towards applying western, high income country, heuristics like left-right dynamics, etc. to the issue as whole. Even countries in the same region have vastly different political setups and dynamics. Focus on a handful of countries, and/or make generalizations when you feel highly confident that they hold in most cases. * **There are many theories of about how policy change happens.** * They often provide useful heuristics to try to understand the system. The following summaries from Paul Cairney are useful: * [Multiple streams](https://paulcairney.wordpress.com/2013/10/31/policy-concepts-in-1000-words-multiple-streams-analysis/) * [Punctuated equilibrium](https://paulcairney.wordpress.com/2013/10/29/policy-concepts-in-1000-words-punctuated-equilibrium-theory/) * [Advocacy Coalition Framework](https://paulcairney.wordpress.com/2013/10/30/policy-concepts-in-1000-words-the-advocacy-coalition-framework/)
'Market shaping' ideas * **Core cruxes to investigate** * Is there a market failure to address? What is it? * We need to be able to point to a market failure with some confidence, and suggest how it is manifesting. We are very unlikely to recommend a market shaping idea without having a good understanding of the market failure it is supposed to address. * Usually, this will require validation through expert interviews. * What are the barriers and catalysts (i.e., leverage points!) which we can engage with to help correct the market? ([see figure 1, page 13 for suggestions](https://www.clintonhealthaccess.org/wp-content/uploads/2026/03/CHAI-Market-Shaping-Guidebook.pdf)) * What are the most likely market shaping mechanisms to be deployed? (a good list of mechanisms can be found in the tables between [pages 14 and 18 here](https://www.clintonhealthaccess.org/wp-content/uploads/2026/03/CHAI-Market-Shaping-Guidebook.pdf)) * Because market shaping mechanisms need to match the market failure dynamics very well, we expect a lot of the detail here to be added by founders themselves during exploration. * However, we should attempt to narrow the field and give guidance on what we think is most likely to work. * **Key guidance resources** * [Clinton Health Access Initiative Playbook](https://www.clintonhealthaccess.org/wp-content/uploads/2026/03/CHAI-Market-Shaping-Guidebook.pdf) * [Renaissance Philanthropy Market Shaping Playbook](https://www.renaissancephilanthropy.org/playbooks/market-shaping)
'Information and evidence' ideas * **Key cruxes** * Determine what the data gap looks like, and why it exists (e.g., it is too expensive to generate the information, there is little interest from stakeholders, etc.) * Build a strong picture of what the use case for the data would be, and validate it. * For example, if you expect a certain type of philanthropic funder to make use of data or evidence that an organization would generate, you should discuss what that use case would look like with those stakeholders. * A key question to answer here is what the next best action would be in the absence of that data or evidence. That helps establish a counterfactual. * Information and evidence ideas often follow a similar pattern to market shaping ideas. For instance, CHAI's [playbook](https://www.clintonhealthaccess.org/wp-content/uploads/2026/03/CHAI-Market-Shaping-Guidebook.pdf) details plays such as "demand visibility", and "forecasting and quantification", which fall neatly into our 'information and evidence' mechanism. Therefore, much of the same advice applies.
'Tech development and adoption' ideas * **Key cruxes** * If the product exists: Why are people not using it already? How and when would/do people use it? * If the product does not exist: Why does this product not exist already? How technologically feasible is it to create this product? What do technical experts think are the key issues in development? * What does the pipeline of new products in this space look like? Why would a new effort be additional? * What are the financial incentives at play here? * Consider thinking of the product through the [target product profile framework](#user-content-fn-2)[^2]. Resources: * [Renaissance Philanthropy Target product profiles Playbook ](https://www.renaissancephilanthropy.org/playbooks/target-product-profile) * [BARDA TPPs](https://medicalcountermeasures.gov/barda/tpp) * Innovate Animal Ag on [in-ovo sexing](https://innovateanimalag.org/egg-sexing)
[^1]: By attribution, we mean the contribution of an intervention to the outcome, separate from other efforts that may be present in the space. [^2]: You don't have to follow all steps in these playbooks, but they can helpful to think of how to approach the research. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://ambitious-impact.gitbook.io/ambitious-impact-docs/explainers-and-house-views/our-charity-taxonomy/guidance-to-adjust-research-for-system-level-interventions.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.