How to Write a Policy Analysis Proposal

A policy analysis proposal sets out how a decision maker will compare responses to a public problem. It should identify the problem, the authority able to act, feasible alternatives, common criteria, evidence to be gathered, affected people, and the limits of the analysis. The proposal is a plan for a defensible recommendation, not a premature declaration that one option is best. Policies and legal powers vary by place and time, so a real proposal must ground its scope in current sources for the relevant jurisdiction.

Define the public problem precisely

Describe who is affected, where, over what period, and what outcome matters. “Transport is inadequate” is too broad; missed access to an essential service in a particular district can be investigated. Use counts and rates with appropriate denominators, and distinguish a rising problem from improved detection. Ask how people who are not represented in official records experience the issue. A vivid anecdote can show a pathway but cannot estimate the burden by itself.

Separate the problem from a proposed solution. If the brief starts by asking for a new bus route, test whether route coverage, cost, scheduling, accessibility, or another factor is the binding constraint. Define a baseline and a plausible causal pathway. A solution chosen before the mechanism is examined may use resources while leaving the most affected people outside it.

Name the decision maker and scope of authority

Identify the institution that can change a rule, service, funding stream, or enforcement practice. A city may control routes but not a neighboring agency’s schedule; a school board may control timetables but not household income. State the legal and budgetary constraints that require verification, and the partners whose cooperation would be needed. Do not write a recommendation for a body that cannot implement it.

Set the time horizon. An immediate operational adjustment and a long-term infrastructure investment may both be relevant but should not be compared as if they have the same delivery period. Define the decision the analysis will support: choose a pilot, allocate a budget, revise a rule, or commission more evidence. This keeps the proposal proportionate to available data and authority.

Construct comparable alternatives

Include a clear baseline and at least two feasible policy options. Describe each with enough detail to estimate its mechanism, cost, and implementation requirements. An alternative should not be a straw man; if it is plainly impossible, replace it with a credible one. Options can combine tools, but identify which component addresses which barrier so their effects can be evaluated.

For a district with missed clinic visits, possible choices might include a shuttle timed to appointments, flexible clinic hours, or community-based follow-up. Each addresses a different part of the journey. The proposal should specify which patient groups can use each, what capacity is required, and whether the option shifts cost to families or providers. A reminder message is a useful comparator only if lack of information is a plausible barrier.

Set criteria before judging options

Use the same criteria for every alternative: expected effectiveness, equity, cost, feasibility, speed, sustainability, and unintended effects as appropriate. Define how each will be measured. “Equity” may mean reach among people facing the greatest barrier; “feasibility” may require staff, vehicles, contracts, and authority by a certain date. Weighting criteria can help, but do not hide a value judgment inside unexplained numbers. Show which trade-off drives the choice.

Gather evidence that fits each criterion. Administrative data can describe use; interviews can reveal barriers; budgets can estimate direct costs; evaluations elsewhere can suggest likely effects with limits of transfer. State assumptions and scenarios when exact estimates are unavailable. A proposal should identify where evidence will come from and how long collection will take, not promise a precise benefit without a method.

Plan engagement and implementation evidence

Affected people should help define the problem and test options. Explain who will be invited, how participation will be accessible, and how their input will change the analysis. A single meeting with a convenient group may miss people who work at the meeting time or cannot travel. Protect confidentiality when the topic is sensitive. Stakeholders with implementation duties should also review capacity and operational constraints.

Plan how each option would work in practice. Who enrolls or qualifies, who communicates the change, how exceptions are handled, and what happens if demand exceeds capacity? An intervention that appears affordable on paper may require an unbudgeted coordinator or a data-sharing agreement. The proposal need not solve every detail, but it must flag those that could reverse the recommendation.

Outline the analysis and decision product

Specify the data collection, comparison period, evaluation method, schedule, and deliverable. A decision table can show the options against criteria; prose should explain uncertainty and trade-offs. Identify a sensitivity analysis for assumptions that matter most, such as uptake or unit cost. A pilot may be appropriate when a full rollout has irreversible cost or uncertain reach. Define what pilot result would justify scaling, revising, or stopping.

The final proposal should state what it can conclude and what it cannot. It should not claim that a survey of interested users represents everyone affected or that a difference after implementation proves causation. End with a timetable for a recommendation and a route for review if circumstances change. A strong policy analysis proposal makes a public choice inspectable before resources and authority are committed.

Before submission, check that every option has been evaluated on every criterion and that the evidence plan does not favor the option the writer already likes. If a criterion cannot be measured within the available time, explain the proxy or qualitative judgment that will be used and its weakness. The decision maker needs to know which uncertainty is consequential, not merely that all forecasts are uncertain.