I don’t know how many opinion pieces I’ve read blaming New Zealand’s ‘tall poppy syndrome’ for weak innovation and productivity growth. But how could you test whether it is true?
In principle, a culture that celebrates achievement will encourage more of it, and a culture that punishes innovators and problematises success will encourage people to leave. Unfortunately, in the absence of reliable data, the talk has largely been waffle.
If you want to test it rigorously, you need to be able to define it, measure it, show how it changes over time, and link those changes in culture to changes in the other things people care about.
It has not been easy.
But we live in an age of wonders. Many things can now be done that would have been impossibly expensive to try even a decade ago.
For example, economists can now show that changes in culture, as measured by themes in a large sample of the Western written canon over the past thousand years, seem to track changes in productivity.
Amazing.
Project Gutenberg is a repository of public domain books. It’s a superb resource for anyone looking for a classic book to read. Unfortunately, excessively stringent copyright rules mean the past century’s books are largely excluded.
University of St Andrews economist Radek Stefanski asked Gemini 2.5 to read through and score the cultural themes in over 23,000 books held by Project Gutenberg. His working paper, Cultural Capital and the Productivity of Ideas, was released this month.
Gemini scored whether each book encouraged readers to respect hierarchy, order and authority, or whether it celebrated autonomy and dissent. Did each one’s themes highlight fear of the unknown, or the wonder of experimentation? And did the book inculcate the values of patient accumulation on which long-term progress is built?
The work leans on contextual assessment that has only recently become reasonably reliable in automated systems, cross-checked against simpler and more standard text-counting methods.
A sample of Gemini’s scoring was checked against DeepSeek v4 (a different AI) and against literary scholars, who gave similar scores for books they knew. And a separate archive of 5,000 books from other collections, scored the same way, showed the same change in culture.
The measures, running for almost a thousand years for the main sample, show trends in elite culture – the culture of those who write, teach, and purchase books. In some periods, elite culture preferred hierarchy and rule-following while fearing the untried. In others, themes more conducive to innovation, like autonomy and patience, dominated.
The work shows that Western culture shifted in more innovation-friendly directions from about 1100AD to the 1300s, and then again from the 1500s onward. A smaller but suggestive set of works showed greater elite support for the conditions for innovation prior to around 500AD, and worsening conditions through the Dark Ages.
Overall, the work suggests that cultural barriers to innovation halved from 1000AD to 1920.
Geographically, Europe’s centre of cultural support for innovation moved northwest from the Mediterranean in 500AD to northwestern Europe by 1900. In North America, the strongest cultural support for innovation moved west as America grew from 1700 to 1900.
Stefanski estimated the effect of cultural change on the number of notable innovators each country produced a generation after its culture shifted. He then fed that relationship into a standard economic growth model.
On that accounting, a 1920 with the culture of 1000 would have been roughly half as productive. We would all be so much poorer.
Greater cultural appreciation for innovation also predicts about two thirds of the first real sustained acceleration of productivity growth from 1500 through 1700. And places with a culture more friendly to innovation showed some evidence of stronger urbanisation over the subsequent century.
The work includes a smattering of older New Zealand books. But if there were any interesting differences between the Antipodes and the rest of the world, they were not reported in the paper.
It suggests a project for a keen local graduate student.
The paper’s method could be replicated across historic New Zealand newspapers and magazines, starting with the Papers Past archive, to test whether the ‘tall poppy syndrome’ really is a thing. Is an innovation-friendly culture a precursor to more innovation and growth? Or do periods of stronger growth build a culture that is more friendly to innovation?
I would love to know the answer. I suspect that cultures that welcome innovation, and that do not punish or berate innovators, matter considerably – and that they matter here as well. But I would rather know than suspect. On this, better-than-waffle is finally possible.
Dr Eric Crampton is Chief Economist at the New Zealand Initiative. This article was sourced HERE
If you want to test it rigorously, you need to be able to define it, measure it, show how it changes over time, and link those changes in culture to changes in the other things people care about.
It has not been easy.
But we live in an age of wonders. Many things can now be done that would have been impossibly expensive to try even a decade ago.
For example, economists can now show that changes in culture, as measured by themes in a large sample of the Western written canon over the past thousand years, seem to track changes in productivity.
Amazing.
Project Gutenberg is a repository of public domain books. It’s a superb resource for anyone looking for a classic book to read. Unfortunately, excessively stringent copyright rules mean the past century’s books are largely excluded.
University of St Andrews economist Radek Stefanski asked Gemini 2.5 to read through and score the cultural themes in over 23,000 books held by Project Gutenberg. His working paper, Cultural Capital and the Productivity of Ideas, was released this month.
Gemini scored whether each book encouraged readers to respect hierarchy, order and authority, or whether it celebrated autonomy and dissent. Did each one’s themes highlight fear of the unknown, or the wonder of experimentation? And did the book inculcate the values of patient accumulation on which long-term progress is built?
The work leans on contextual assessment that has only recently become reasonably reliable in automated systems, cross-checked against simpler and more standard text-counting methods.
A sample of Gemini’s scoring was checked against DeepSeek v4 (a different AI) and against literary scholars, who gave similar scores for books they knew. And a separate archive of 5,000 books from other collections, scored the same way, showed the same change in culture.
The measures, running for almost a thousand years for the main sample, show trends in elite culture – the culture of those who write, teach, and purchase books. In some periods, elite culture preferred hierarchy and rule-following while fearing the untried. In others, themes more conducive to innovation, like autonomy and patience, dominated.
The work shows that Western culture shifted in more innovation-friendly directions from about 1100AD to the 1300s, and then again from the 1500s onward. A smaller but suggestive set of works showed greater elite support for the conditions for innovation prior to around 500AD, and worsening conditions through the Dark Ages.
Overall, the work suggests that cultural barriers to innovation halved from 1000AD to 1920.
Geographically, Europe’s centre of cultural support for innovation moved northwest from the Mediterranean in 500AD to northwestern Europe by 1900. In North America, the strongest cultural support for innovation moved west as America grew from 1700 to 1900.
Stefanski estimated the effect of cultural change on the number of notable innovators each country produced a generation after its culture shifted. He then fed that relationship into a standard economic growth model.
On that accounting, a 1920 with the culture of 1000 would have been roughly half as productive. We would all be so much poorer.
Greater cultural appreciation for innovation also predicts about two thirds of the first real sustained acceleration of productivity growth from 1500 through 1700. And places with a culture more friendly to innovation showed some evidence of stronger urbanisation over the subsequent century.
The work includes a smattering of older New Zealand books. But if there were any interesting differences between the Antipodes and the rest of the world, they were not reported in the paper.
It suggests a project for a keen local graduate student.
The paper’s method could be replicated across historic New Zealand newspapers and magazines, starting with the Papers Past archive, to test whether the ‘tall poppy syndrome’ really is a thing. Is an innovation-friendly culture a precursor to more innovation and growth? Or do periods of stronger growth build a culture that is more friendly to innovation?
I would love to know the answer. I suspect that cultures that welcome innovation, and that do not punish or berate innovators, matter considerably – and that they matter here as well. But I would rather know than suspect. On this, better-than-waffle is finally possible.
Dr Eric Crampton is Chief Economist at the New Zealand Initiative. This article was sourced HERE
No comments:
Post a Comment
Thank you for joining the discussion. Breaking Views welcomes respectful contributions that enrich the debate. Please ensure your comments are not defamatory, derogatory or disruptive. We appreciate your cooperation.