How can I compare candidates without a popularity contest?

A useful comparison starts with the office being contested. Someone standing for Parliament should not be assessed as though they directly operate every municipal service. Learn the role, then decide which responsibilities matter most to you. 1 2

Pick a small number of questions

Choose issues you can investigate: a policy position, a documented intervention, a disclosed interest or a promise about public spending. Apply the same questions to every candidate you compare.

For a previous office-holder, find the relevant original record rather than relying only on a campaign summary. For a new candidate, do not invent an equivalent performance history. Label the evidence as a proposal, a past record or an unanswered question.

Separate information from your judgment

A comparison can have four columns: claim, source, what it establishes, and your assessment. A fifth column for “unknown” is often more honest than a forced numerical score.

Your priorities are not measurements. You might care more about a transport policy than another reader does. State that choice instead of pretending a personal weighting is an objective scientific ranking.

Use affidavits carefully

Compare matching fields and dates. Treat an independent extraction as a useful aid, not a reason to ignore the original affidavit. Criminal-case disclosures require particular care: a declared case is not a finding of guilt, and a past election document does not establish present case status. 3 4

Avoid creating a “criminality score” by adding together allegations, charges and convictions. Those categories do not mean the same thing.

Test a promise against powers and costs

In a hypothetical example, two candidates promise more frequent buses. Ask which operator controls the service, what action the elected office could take, what funding is proposed and how delivery would be measured. Prefer a specific answer that acknowledges these responsibilities to an unsupported claim of personal control.

For a spending promise, distinguish the proposed amount from an authorised allocation and from actual expenditure. The official budget documents themselves make stage distinctions; your comparison should not erase them. 5

Leave room to change your view

Record when you checked the material. A corrected source, a new policy document or a current court order can change an assessment. Updating the comparison is better than defending a conclusion after its supporting evidence has changed.