Why public survey on Hong Kong’s long-term development is just mindless number-crunching
The next time the government announces the results of a “public engagement exercise”, gauging citizens’ opinions on a policy proposal, remember this computing term: garbage in, garbage out. That is, flawed input will result in faulty output.
Garbage in, garbage out encapsulates the poor design and sloppy methodology of government-sponsored surveys which produce worthless results.
Take the latest “public engagement” on Hong Kong’s long-term development plan, dubbed 2030+.
Costing HK$1.04 million and conducted by the University of Hong Kong’s Social Sciences Research Centre, the survey used no statistical sampling, which involves selecting a random representative set of people to estimate the responses of the whole population. Without statistical sampling, those polled are not a representative sample of citizens and no valid conclusion can be drawn from the results, as any professional pollster can tell you.
For 2030+, the government solicited a huge quantity of feedback, ranging from comments expressed in large and small groups to answers to questionnaires. Thousands of individuals and organisations also submitted petitions.
Astoundingly, no statistical sampling plan was used in capturing so much data. Feedback was instead divided into subsets, providing no insight on the aggregated opinion of the population at large. Without a sampling plan, the statistical significance of the population subsets cannot be inferred. Without an analytical framework to organise data and interpret the results, the whole exercise is mindless number-crunching.
More egregious is the way the exercise used leading questions to elicit public approval. People were presented with the 2030+ vision of an ideal city described by feel-good rhetoric – vibrant, healthy, living space, quality jobs and other buzzwords.
Who would object to such a city? The survey then asked about the desirability of colossal housing projects, such as the East Lantau Metropolis (ELM), as if these would realise the ideal. By linking the ideal to the proposal, the survey misled people into approving the latter.
No mention was made that ELM will cost HK$400 billion, not be ready for decades and so offer no solution to the current housing crisis. Instead, if constructed, it will result in future overcapacity leading to housing a population of 9 million in Hong Kong when the peak is 8.22 million in 2043, decreasing to 7.8 million in 2064.
Does the government intend to make policy based on such worthless research? Or perhaps it never intended to gauge public opinion, merely to give the appearance of doing so.
Tom Yam, Lantau