Skip to main content
Explanation

How to evaluate a proposal

Political discussions often mix five types of claim. Separating them improves the conversation: value, observed fact, causal explanation, hypothesis and choice.

Five types of claim

The examples below illustrate types of claim. They are not observed results from specific programmes.

  1. Value

    Example“No one should be abandoned because they are unable to work.”

    How to examine itProvide a moral justification and discuss limits and conflicts.

  2. Observed fact

    Example“This programme served a certain number of people during this period.”

    How to examine itCheck the source, period, coverage and definition.

  3. Causal explanation

    Example“The rule change caused the increase in employment.”

    How to examine itCompare it with what would probably have happened without the change.

  4. Hypothesis

    Example“Portable support may make it easier to change jobs.”

    How to examine itExplain the mechanism and design an evaluation.

  5. Choice

    Example“We will prioritise this policy within this budget.”

    How to examine itShow the alternatives, costs, criteria and who decides.

Eight questions for any proposal

  1. What problem is being solved? A precise description is more useful than a generic enemy.
  2. Who experiences it? Averages can hide differences in location, income, age and need.
  3. What already happens today? The proposal needs to start from existing institutions and rules.
  4. How would the change produce the outcome? A sequence of actions lies between spending and improving people's lives.
  5. What evidence supports this expectation? A success story, an association and a causal evaluation have different strengths.
  6. What is the full cost? Implementation, operation, maintenance, transition and risk need to be shown.
  7. What is the best available alternative? Including improving what already exists.
  8. What would lead us to correct or end the intervention? The answer should be formulated before the outcome is known.

This topic in the model

Comparing projects is one stage of OpenGov3's proposed election and governance model. Objectives, alternatives, costs and assessment criteria would need to be clear before the choice.