You are subscribed as | Unsubscribe | View online version | Forward to a friend

Insights 39 : 16 October 2020
Newsroom: Eric Crampton says Green wealth tax will impact many more people than they think.
Podcast: Briar Lipson responds to criticism about her education report.
Report: Roadmap for recovery: Briefing to the incoming government.

Effective duty
Eric Crampton | Chief Economist |
People are more careful when spending their own money than when spending on behalf of others. This truism is especially relevant in election years, but it applies more broadly.

Harvard and Massachusetts Institute of Technology researchers were keen to understand why people care more about the effectiveness of their spending when that spending helps themselves and their families, as opposed to helping strangers. The results were published earlier this week in Nature: Human Behaviour.

The researchers gave experimental subjects some money and an opportunity to multiply the amount by saving the money, giving the money to a family member, or by giving the money to charities or strangers.

The recipients were keen to put the money into savings accounts, or to the benefit of their kin, when the donated amount was multiplied by the researcher. The effectiveness of giving money to one’s future self or to one’s family mattered. But not so much when donating to charities or strangers. Donations to others were about the same regardless of any multiplier.

The result may be depressing, but not particularly surprising. Yet the reason was intriguing.

In another round of experiments, donors were rewarded by another experimental subject: an observer who saw the amount donated, and if the experimenter multiplied the donation. Observers rewarded donors for their donation amount, but not for its effectiveness. In a separate survey, observers’ judgements of the donor’s character also varied only by the donation’s amount, not by the donation’s effectiveness.

To put it simply, people are more wired in social settings to care about being seen to do good than whether their actions are truly helpful. Effort earns a social reward, not the effectiveness.
Successful electoral campaigns play to those hard-to-overcome features of human nature. In a better world, charitable donations would be driven by effective altruism, and voting would be based on which policies do the most good towards the ends folk most care about.

The lesson? Make every ballot an effective one. Weigh up which promised policies have the best chance of improving outcomes, or at least not doing harm. The policies offered by any party will not substantially improve until voters start rewarding effectiveness rather than intentions.

The (education) empire strikes back
Roger Partridge | Chairman |
Last week, the Initiative took aim at those presiding over the country’s education system. For nearly two decades Kiwi students have suffered a steady decline in performance. Where once our school system was the envy of the world, now it is barely mediocre.

The report, New Zealand’s Education Delusion: How bad ideas ruined a once world-leading school system, places the blame squarely on the education establishment. Author Briar Lipson writes there is a “rot at the core of schooling in New Zealand. The Ministry of Education follows unscientific advice and is in thrall to a flawed philosophy.”

Called “child-centred learning,” this philosophy places children at the centre of decisions about their education. This may sound innocuous. But a combination of factors has made it a damaging dogma. What a child now learns depends on the individual school, teacher and, above all, child. The outcome is that education has become a lottery. Little wonder standards have slipped.

Predictably, the report prompted a fierce counterattack from the education establishment. Leading the charge is Claire Amos, outspoken advocate of the child-centred approach and principal of Albany Senior High School.

Writing in The Spinoff last week, Amos’s assertions were preposterous and replete with personal attacks.

The Initiative’s report cites empirical studies and cognitive science. Amos describes these as “hyperbolic opinion.” For Amos, that Lipson was associated with a Conservative think tank in the UK discredits the report’s evidence. Yet the science is apolitical. That, after all, is the hallmark of science.

More troublingly, Amos sees the call in New Zealand’s Education Delusion for a knowledge-based curriculum as a “colonial tool putting old Western knowledge ahead of indigenous communities.” On similar grounds, she dismisses Kiwi students’ two-decade long decline in the OECD’s PISA results in reading, maths and science.

Yet there is nothing Western about wanting every child to read, do basic maths and have a workable understanding of science. Education is as important to a child in Karachi, as it is to a child in Khandallah or Kaitaia.

Children deserve to attend schools that equip them with the best education to fulfil their potential. This is not Western. It is a matter of social justice.

The evidence in New Zealand’s Education Delusion should be enough to persuade any open-minded educator to question the system’s dogged adherence to the child-centred approach. Two decades of damage to Kiwi school children is surely enough.

The last human prime minister
Nathan Smith | Chief Editor |
A calendar app on Jacinda Ardern’s monitor shows the year is 2023. Outside is cold and threatening rain - another day in Wellington.

Ardern leans back in her brown leather chair and squeezes the lid closed on the ballpoint pen, thinking through yesterday’s meeting with a deputy secretary about the wonderful success in halving the housing wait list – a key Labour promise in the 2020 election.

She was happy and wanted to know how the computer models helped achieve this. It wasn’t obvious when they looked at it, the secretary said, so they would manually review it and come back to her.

Ardern couldn’t shake a strange feeling. If anyone knew how the models worked, it’d be the secretary. And if he couldn’t say, then who was really giving her the report? Was it the secretary or the computer models?

The prime minister wondered how many reports over the last three years actually involved human beings who really understood how artificial intelligence worked.

The housing outcome was good news, sure. But what if the computers slowly began to flag a slightly different set of data about housing? Would the programmers catch the subtle shift?

Apparently not, because it was a bit too complex even for the secretary. He mentioned the people who know about housing aren’t the same people who programmed the algorithm. Also, it’s not like there was one programmer. A team wrote the code, each working on only a small chunk.

Maybe it was the weather, but a thought popped into Ardern’s mind: does the complexity of this AI give it free will?

The raindrops running down the windowpane don’t have free will. But human free will is based on the fact that you can’t predict what I’ll do. Brains are too complicated. Yet if you knew every input and variable going into my head, you could probably predict my behaviour.

After all, the only thing that gave Ardern the impression of free will is that even she didn’t know what she was going to do sometimes.

Staring out the 9th floor window, Ardern realised she didn’t need to hear the secretary’s answer. If the algorithm was too complex to understand, something a lot bigger than the weather had just changed.

Today it was confirmed that the government’s AI had something like free will. It was also the last day that a human prime minister was fully in charge of New Zealand.

On The Record
All Things Considered
Copyright © 2024 The New Zealand Initiative, All Rights Reserved

Unsubscribe me please

Brought to you by outreachcrm