4 Evaluating Policy Impacts and Trade-Offs
Learn how to compare policy options against a baseline, assess their impacts across groups and over time, and communicate uncertainty and trade-offs clearly.
What a policy assessment should show
A policy assessment should consider more than whether an option addresses a problem. It should make visible the likely effects, who experiences them, what they cost, and how uncertain the estimates are. This helps decision-makers weigh competing goals transparently.
Compare options with a
Compare every option with a clear : what is likely to happen without the proposed policy, including relevant existing programs and expected trends. Estimate effects relative to that over a defined period.
Consider practical alternatives, including non-regulatory approaches and the status quo, so that the comparison is not tilted toward a predetermined solution. Include direct and indirect effects, intended and unintended consequences, and implementation and compliance costs.
For each option, assess:
Benefits: improvements in outcomes, such as better health, safety, access, or environmental quality.
Costs: resources used by government, affected organizations, households, or others. Include administrative, implementation, and compliance costs where relevant.
Other effects: consequences that matter to the decision but are difficult to quantify or express in money.
Estimate effects as carefully as the available evidence allows. compares benefits and costs in monetary terms when suitable. compares the cost of achieving a specified outcome. Neither approach removes the need to describe important impacts that are not monetized.
Identify who gains and who bears costs
An overall positive result can conceal uneven effects. asks how benefits and costs differ across groups, places, and time—for example, by income, age, geography, or business size. The people who bear a policy’s costs may not be the people who receive its benefits.
Report significant differences separately from economy-wide totals, and identify groups facing especially large or severe effects.
For example, a city considering heat protections might compare cooling-center access, home-weatherization grants, and a building cooling standard. Cooling centers may offer quick help but be less accessible to people with limited mobility or transportation. Grants may reduce household costs but require funding and an application process. A standard may improve protection broadly while imposing upfront costs on building owners, with some costs potentially passed on to tenants. These are questions to investigate, not guaranteed outcomes: local data and policy design determine the actual effects.
Make uncertainty explicit
Forecasts depend on assumptions about participation, costs, behavior, and future conditions. Identify the main uncertainties and test whether conclusions change when key assumptions vary.
changes one or more important inputs. compares coherent sets of assumptions, such as low, central, and high participation. Present ranges or alternative results where appropriate, and distinguish well-supported estimates from speculative ones.
Uncertainty alone is not a reason to omit a potentially important effect. Instead, explain the evidence and its limits carefully.
Explain trade-offs across goals and time
Options can perform differently across goals: one may produce greater benefits but cost more, act more slowly, or place greater burdens on a particular group. Compare options using consistent assumptions and criteria.
If benefits and costs occur at different times, show their timing. Monetary comparisons across years may use ; disclose the rate and test how it affects the result. When effects cannot reasonably be reduced to a single monetary measure, present them alongside quantified results or use explicit criteria to compare them.
Do not let a single total hide important harms, distributional consequences, or uncertainty. State why an option is preferred, which trade-offs remain, and what evidence could change that judgment.
Plan to monitor results
Identify indicators and data needed to monitor results after implementation. Monitoring makes it possible to compare forecasts with observed outcomes and to assess whether new evidence could change the original judgment.
The aim is not to make uncertainty disappear. It is to make the evidence, its limits, and the value judgments clear enough to support informed decisions.