Labour dominates Ipsos poll, but rival survey raises questions about online-only sampling

Summarised by Centrist

Labour is rated better than National at managing New Zealanders' five biggest concerns in the latest Ipsos Issues Monitor, with the opposition leading on 16 of 20 issues surveyed.

However, an earlier October survey using a different methodology gives National the lead on seven of ten measures, raising questions about how polling methods shape the picture presented to voters.

The latest Ipsos survey, reported by RNZ and the NZ Herald, identifies inflation and the cost of living as the biggest concern, cited by 64% of respondents.

Healthcare follows at 45%, with fuel prices jumping 14 percentage points to 31%, overtaking the economy at 30%. Unemployment ranks fifth at 19%.

Labour leads National on cost of living by 38% to 24%, healthcare by 42% to 21%, and fuel prices by 35% to 20%.

National retains a narrow lead on crime and law and order, 29% to 28%, its only lead among the 20 issues measured.

The government's overall performance rating remains unchanged at 4.1 out of 10.

Ipsos surveyed 1,008 adults between 19 and 27 October using online research panels.

The results measure public perceptions of party capability, rather than independently assessed government performance or voting intentions.

Editor's note: Monday's Three News broadcast highlighted Labour's lead on all five major voter concerns in the Ipsos survey. Although its graphic identified the survey as an online sample, the broadcast did not compare the findings with alternative polling.

In the publicly released October Taxpayers' Union-Curia survey, National led on seven of ten measures including the economy, inflation, education, safety, spending, the environment and not increasing taxes. Labour led on health, poverty and housing.

The contrast with Ipsos is substantial, particularly on inflation, the economy and education.

One important methodological difference is that Ipsos relies entirely on online panels, while Curia interviewed 700 people by telephone and 300 online.

Online panels can be vulnerable to selection and coverage biases. Weighting results by age, gender and region helps correct demographic differences but cannot necessarily account for differences in political attitudes among people who participate.

Ipsos has itself acknowledged these limitations in its general research guidance. Curia also uses online respondents, and its earlier fieldwork, focus on likely voters and differently worded questions complicate direct comparisons.

Market analyst Geoff Neal, co-founder of KPI.nz, argues that perceptions should be tested against measurable outcomes in areas such as inflation, housing, healthcare and crime, rather than treated as evidence of governing performance.

For example, despite Labour's substantial lead on healthcare in the Ipsos poll, Health New Zealand reports year-on-year improvements across all five government health targets. Those improvements do not establish which party deserves credit, but provide additional context.

Read more at RNZ

Re-published from the Centrist with permission

Original article at The Centrist