The New Zealand Paradox
What can we do to lift prosperity in New Zealand? The tools and concepts needed to answer this question have their origins in statistical physics. This is a problem tailor made for physicists and mathematicians at Industrial Research Ltd. Professor Shaun Hendy explains.
Nanotechnologists sometimes trace the origins of their field back to a 1959 lecture by Richard Feynman, “There’s Plenty of Room at the Bottom “. Feynman foresaw the day when it would be possible to manipulate individual atoms to make new devices and technologies. Scientists at the MacDiarmid Institute now do this, if not routinely, at least often enough that they don’t feel the need to issue a press release or run down the corridor yelling “Eureka!”.
Today, there are those that would label this type of science as reductionist in a certain tone of voice. Fine, nanotechnology is that and more, but in such circles reductionism has become a pejorative. There is a sense that scientists are missing something when they break matter down into its constituent parts, that somehow the whole is greater than the sum of the parts.
It’s not just trendy French philosophers that are skeptical. Much as Feynman foresaw nanotechnology, disdain for reductionism was anticipated by celebrated theoretical physicist, Philip W. Anderson. While Anderson is perhaps best known for his work in superconductivity, in 1972 he published an article in Science called More is Different arguing that being a reductionist does not make you constructionist.
Anderson saw that even as we deepen our knowledge of atoms and their electrons, matter will still surprise us with phenomena like superconductivity. We may be able to explain such things after we’ve found them in the lab, but complex collective phenomena are almost impossible to predict before they have been seen. In his article, Anderson focuses on how the fundamental physical laws that are obeyed by individual atoms can be broken as a collection of atoms get larger and larger.
Strangely enough, forty years on, New Zealand’s economy faces exactly this sort of problem. Economists have even given it a name: the New Zealand paradox. New Zealand’s recipe for economic growth based solely on efficiency, markets and monetary policy has not enabled us to catch the rest of the developed world. The missing ingredient has been innovation, and when it comes to innovation one thing is becoming clear: more is different.
This is the problem my new team is trying to tackle at Industrial Research Ltd. Understanding how the economic decisions of individuals and firms combine to drive innovation and economic growth is not a solved problem in economics. So Dion O’Neale, Gregor Neumayr and Philip Zhang have joined me to apply methods developed in statistical physics to help understand how New Zealand can increase innovation and economic growth. As Anderson observed, it is almost impossible to predict all the properties of complex systems from first principles, so our approach has been heavily driven by data.
At the heart of macroeconomics are things called aggregate production functions; these are the black boxes used by economists to model the output of an economy. Unlike thermodynamics, say, the laws of macroeconomics have not yet been related to an underlying microscopic economic theory and so at best aggregate production functions subdivide the economy into only a handful of sectors, often into something like traditional and modern.
Can such an approach work? Although I am the first to appreciate the value of simple models, aggregating production in this manner ignores Adam Smith’s insight that specialization can drive productivity growth. Modern macroeconomics captures productivity growth by modelling the invention of indistinguishable. Why does New Zealand export quartz crystal oscillators and but not accelerometers, and sell both sleep apnea devices and respiratory humidifiers? Economics is largely silent on this.
Recently, physicist Cesar Hidalgo and colleagues have put together a map of product space. Their map is built by observing the basket of goods that countries produce and linking together products that are typically found together in the same baskets. They show that some regions of product space are more complex than others, containing many densely interconnected product lines.
These maps provide a way of representing the complexity of a country’s economy. A more complex economy will tend to contain firms that are more specialised, and according to Adam Smith, a more specialised firm is a more productive firm. Indeed, Hidalgo’s analysis shows that countries with more complex economies are richer.
Recent work by my team looks at the complexity of the economy from another angle. Dion O’Neale has been following up on initial investigations of the distribution of patents amongst companies. There are many companies with a few patents and few companies with many patents, and this fall off in patents per company turns out to be well described by a power law.
What Dion has found suggests a convergence of the distribution as countries get richer. Poor countries tend to have a steeper decay in patents per company, while rich countries seem to be converging on a shallower power law. New Zealand is about half-way there, showing that we have room in our economy for IP-rich firms.
The power law exponent correlates well with business expenditure on R&D, up to an intensity of about 1.6% of GDP. Once this level of private sector R&D spending is exceeded, the slope of patent distribution amongst companies converges to a universal value, and further spending just seems to grow the total area under the distribution curve. Although the economy is obviously an incredibly complex system, these universal behaviours that Dion has uncovered are remarkable.
Other work in my team has looked at the interconnection between firms by looking at co-patenting relationships. We can link Fonterra to IRL to F&P Healthcare to AgResearch to Pfizer through relationships between inventors. The web of interconnections is dense and not obviously sectoral based.
What we are seeing is that a productive economy is a complex economy. To build complexity, we will need a more diverse innovation ecosystem. We will need to spend more on R&D and resist the urge to pick winners.