Archive for May, 2019

Please use the format “Very few X are Y” rather than “Not all X are Y”

May 31, 2019
Once again someone in the media uses “Not all X are Y… ” construction.
I would like to see it called out, and have those calling it out formally ask that people who use the structure rephrase it as – for examples – “Very few men commit murders, but most murders of women are by men;” “Very few men are rapists, but most rapists are men;” “Very few (insert religion here) are terrorists, and most religions have spawned a few terrorists;” and “Very few people drive while drunk.”
– because humans tend to do what they think the majority of their group do, so it is wise to emphasise that wrongdoers are a minority in a group:  that way the likelihood of modeling bad behaviour for that group’s members reduces.
If you don’t believe me, check the research on “nudging.”

Come to the party? I want to start the Best Evidence Party.

May 8, 2019

After spending 3 hours researching the parties vying for positions in the Oz elections, I find we need another party.  I want to start it.


Because in the Senate vote, a valid vote must nominate at least 6 parties (above the line vote) or 12 individuals.  There is a huge list of parties, but in my State there are only 5 parties for which I could consider voting – and that is only because I include the far left to balance out the media support for the far right, and also include special interest parties with limited ranges of policies.

Few parties base any of their policies on best advice from the majority of acknowledged international experts – and that’s not a uniquely Australian problem:  I remember hearing of a senior USA economist being happy because he got their government to shift from something like the 17th worst option to the 14th.  Most of those that claim they have based policies on evidence rely on cherry-picked, biased reviews of serious research, seriously flawed research, accumulated anecdotes, their memories of what they learned in high school, or their imaginations. This annoys me.

The vast majority of minor parties show policies which are swayed more by the mass media than by serious study of the complexity of global environmental, trade, economic, and legal systems.  Their policies are usually based more on the religion and customs of their upbringing rather than on serious study of the wide range of predictable cultural and interpersonal differences which make up the range of our citizens’ needs.  This annoys me.

It is time for a party where the overarching policy is “to weigh all proposed legislation in the balance of the best available evidence.”

I feel that a small range of formal policies is better: explaining the approach is better than arguing details of cases, when every scientist knows that we must shift our understanding when further evidence conflicts with what we thought was the case.

I would suggest that its policies would include things like

  • Have the National Broadcaster allocate time to disagreeing speakers, documentaries, etc, on topics where expert evidence is available in proportion to the depth of evidence on their side.  If only one in a hundred internationally accredited experts agree with a view, it should get a hundredth of the time, not equal time.
  • Our Members of Parliament will not promise to reflect the opinions of the electorate.  They will promise to do their best to weigh the evidence, including details not generally known in the electorate, and to consider advice from the wisest advisors available.  This will be the basis of their voting in Parliament.
  • Where research is cited concerning a view we are asked to support, our party will consider advice from experts including experts in the mathematics of statistical analysis:  in a “scientific” culture where a peer-reviewed publication includes a claim that we should see a correlation of 0.1 as “significant”, peer-reviewed publication does not equal depth of evidence.
  • Where reliable evidence is not available, our party would give weight to ideas of responsibilities which have been seen in the best societies and many religions.  Ideas such as: the responsibility of the government to take action to ensure reasonable quality of life for the citizens (Not necessarily paid employment: should the education system be for “a job” or “to learn what you need to be able to learn what you need to have a good life”?); the responsibility of the  top 20% to support the disadvantaged; the responsibility of each to contribute to the well-being of all others in the population; and our responsibility to limit our consumption as required to ensure the chance for following generations to have a reasonable quality of life (We might agree that non-renewable resources should be recycled as far as possible, for example, and argue that one-off or mining profits must not be spent on recurrent expenditure or tax cuts.)

Its focus would not be science, but its approach would often be scientific.  Its considerations would include the long-term consequences of actions, thinking in terms of hundreds of years.  It would attract people who might also consider the Pirate Party or the Science Party, but want a different (or smaller) range of policies.

In Australia, an official “Party” needs 500 voters who are not members of another political party, and a written Constitution.

Would you like to join this party?  Would you like to help draft its constitution?