Is Australia complex enough?

Specialisation as an instrument of economic growth is close to an undisputable fact.  Whether there are limitations to this process, I am not sure, but for me there is a more significant question that really needs a satisfactory answer if we are to understand economic growth – if specialisation causes economic growth, what causes specialisation?

Intuitively is appears that knowledge is the driver of specialisation and our ability to expand the production frontier.  It is new knowledge, especially in the realm of scientific understanding of the material world that enables us to use natural resources more productively, while knowledge of social systems and psychology enables us to share productive gains effectively.

Knowledge as a means to growth is an idea being investigated by a group of physicists-turned-economists at Harvard who have mapped the complexity national economies in an effort to gain an understanding of the relationship between knowledge, and the way knowledge is shared, and economic growth.  In their Atlas of Economic Complexity they explain the transmission mechanism between knowledge and economic growth using the ideas of complexity.

… products are made with knowledge. Consider toothpaste. Is toothpaste just some paste in a tube? Or do the paste and the tube allow us to access knowledge about the properties of sodium fluoride on teeth and about how to achieve its synthesis? The true value of a tube of toothpaste, in other words, is that it manifests knowledge about the chemicals that facilitate brushing, and that kill the germs that cause bad breath, cavities and gum disease.

They go further to explain how it is the way specialised knowledge gained by individuals, then shared throughout society, that promotes both complexity and growth.

We rely on dentists, plumbers, lawyers, meteorologists and car mechanics to sustain our standard of living, because few of us know how to fill cavities, repair leaks, write contracts, predict the weather or fix our cars. Many of us, however, can get our cavities filled, our cars repaired and our weather predicted. Markets and organizations allow the knowledge that is held by few to reach many. In other words, they make us collectively wiser.

The amount of knowledge embedded in a society, however, does not depend mainly on how much knowledge each individual holds. It depends, instead, on the diversity of knowledge across individuals and on their ability to combine this knowledge, and make use of it, through complex webs of interaction. A hunter-gatherer in the Arctic must know a lot of things to survive. Without the knowledge embedded in an Inuit, most of us would die in the Arctic, as has been demonstrated by the number of Westerners who have tried and failed. Yet, the total amount of knowledge embedded in a hunter-gatherer society is not very different from that which is embedded in each one of its members. The secret of modern societies is not that each person holds much more productive knowledge than those in a more traditional society. The secret to modernity is that we collectively use large volumes of knowledge, while each one of us holds only

a few bits of it. Society functions because its members form webs that allow them to specialize and share their knowledge with others.

Analysis of complexity requires no assumptions about marginal costs, supply and demand curves, nor any other of the neoclassical dogma that espouses how society should function to maximise productive capacity.   The linkage from knowledge, to complexity, to economic growth, occurs under any political system and method of trade.  As long as knowledge is accumulated and shared effectively, the result will be continued prosperity.

The plot below shows a global comparison of the relationship between estimated economic complexity and GDP per capita.  The red points are countries that have natural resource exports greater than 10% of GDP, while the blue points show the remainder of countries.  The research clearly reveals that countries with strong reliance on natural resource exports have higher GDPs per capita for a given level of complexity than less resource intensive economies.  Australia did not qualify for as a high natural resource exporting nation in 2009, in part because our total exports are just 21% of GDP (compared to, say, 50% in resource poor Switzerland), and in part because of the slump in resource prices during 2009, particularly of coal and iron ore.  All other years since 2007 would probably have seen Australia marked in red.

There are, of course, some fundamental limitations to the study.  The first is the fact that it relies on measuring the diversity of goods actually produced as a proxy for knowledge.  This overlooks problems inherent in global financing and currency values, especially the impact of Dutch Disease.  Some industries, such as manufacturing, may relocate to cheaper developing countries if currency pressures render them globally uncompetitive.  However, not all knowledge is lost in this process; at least not to the degree that measured levels of production are lost from the figures.

This study of complexity is also limited by the fact that national boundaries are quite arbitrary in terms of globally complex production systems.  The Airbus A380, for example, is manufactured, in part, in possibly more than 20 countries.  The complexity of production required is such a task is shared amongst many nations, potentially decreasing their measured complexity.  But as a global system complexity has probably increased.  Which leads to the more sensitive question of whether domestic complexity is necessary at all so long as a country has strong trade links to the globally complex economy?

Before looking to the policy implications of this type of research, I want to quickly look at Australia’s relative position in the global complex. Australia ranks 79th on the measure of economic complexity, behind Albania, El Salvador, Guatemala, Kyrgyz Republic, Namibia, Kenya, Senegal and most importantly New Zealand (which is ranked 48).  The top three complex economies are Japan, Germany and Switzerland.

In the measure of change in economic complexity between 1964 and 2008, Australia is ranked 95.  In fact we fell from a rank of 32 to 54 over the period.  Of course, much of this is due to the catch-up growth seen in Asia, the Middle East, and Latin America, and we would expect developed countries to perform poorly on this measure.

For me the most interesting figure is the measure of expected growth, which is derived from the mismatch between the complexity measure, and the GDP measure.

China, India and Thailand are at the top of this ranking, since they are countries with economies that are remarkably complex, given their current level of income, and are expected to catch up faster than other developing nations. Next come Belarus, Moldova, Zimbabwe, Ukraine and Bosnia-

Herzegovina, five countries where the current level of income is dramatically lower than what one would expect given their productive capabilities. This ranking shows that the two regions of the world where the potential of per capita growth is higher are East Asia and Eastern Europe (Map

At the bottom of this ranking we have Sudan, Angola and Mauritania. These are developing countries where the complexity of their economies does not provide a basis for future economic growth, and, where changes in income are dominated by fluctuations in the price and volume of natural resource activities

The map below shows the national complexity rankings.

If the theory of complexity and the empirical measurements of indicators of economic complexity have merit, what policy implications should be considered?  My interpretation is that theories of economic complexity suggest that government needs to be an active player in facilitating a diverse and complex economy – from funding or facilitating infrastructure networks (fibre, ports, roads etc), to fostering education, to investing in new technology start-ups and supporting domestic emerging businesses through government purchasing decisions.  In particular it would appear that governments have a role heavily investing in research, particularly ‘blue-sky’ research which often has no immediate commercial value but can open doors to other discoveries, and facilitating market investment in research.

Economic complexity relationships appear to also support notions of a resource curse.  Australia sits far above the trendline for GDP given our level of complexity, meaning our economic performance is highly dependent on the whims of resource markets.  It also appears we could trade-off some of this resource income for investment in diversity of production and maintain our standard of living in a more stable economy.

This research on complexity shows that while specialisation is effective within countries, specialisation of countries themselves usually inhibits growth.  Some may argue that we should be considering complexity of the world economy as a whole, ignoring any within country declines in complexity, but to me this ignores geo-political realities and the principles of national democracy.

Tips, suggestions, comments and requests to [email protected] + follow me on Twitter @rumplestatskin


  1. do countries start with high complexity and then leverage it to economic growth or is it something that can be created by policy?

    i know Thailand pretty well and it seems to me that their economic complexity has been there for a very long time – everybody works, everybody tries to eke out survival on whatever resources are available in their local environment – and there does not seem to be a lot of policy involved in the underlying culture – it’s quite literally a market economy

    the new economic activity in Thailand – hard disk factories, electronic motor factories etc seem to me to be dragging Thailand towards specialisation and less economic diversity – where the families of employees might have come from dozens of different economic backgrounds the employees are very specialised and do basically one thing

    perhaps then the growth anticipated for Thailand is not so much about economic diversity but cultural attitudes to work?

    the idea that you would live on welfare in Thailand is something most Thais would consider absurd

    the idea that you would form a union to battle your employer would also be considered alien

    so if a population has a very very strong work ethic (no work = no eat, no eat = die or put unfair pressure on family) then leveraging that into low cost labour for hard disk factories seems to be a no-brainer

    and the commensurate rise in GDP would be more about unmeasured productivity being converted into measured productivity

    eg: in rural Thailand everyone has a garden – but the production from the garden is not likely to end up in GDP figures


    • Interesting points pop, although this is the opposite to how complexity is measured in this research.

      “the new economic activity in Thailand – hard disk factories, electronic motor factories etc seem to me to be dragging Thailand towards specialisation and less economic diversity – where the families of employees might have come from dozens of different economic backgrounds the employees are very specialised and do basically one thing perhaps then the growth anticipated for Thailand is not so much about economic diversity but cultural attitudes to work?”

      As the inuit example shows, specialisation increases complexity, rather than decreases it, because it allows a country to produce products other countries can’t. So in Thailand, sacrificing a variety of common industries for a few specialty industries will increase complexity. You don’t get a great complexity measure because you produce lots of different things that many other countries also produce. You get a good measure to being able to do that AND produce things that other countries don’t produce.

      The study uses the term ubiquity as a measure of how many other countries can make the same products. They couple this with a measure of diversity (how many products the country makes) to come up with the measure of complexity (p20)

  2. This is an extremely useful framing of a principle I have held to for a long time. A great study, thanks for putting it up, Rumplestatskin.

    Yes, wealth depends on specialisation and trade. Interestingly, some highly-populated but resource-poor nations, like Japan and the Netherlands, end up almost “obliged” to succeed in “value added” export led growth; while under-populated and “resource-rich” nations seem to rest on their laurels and suffer relative decline over the long term.

    Having to import most of your food and commodities is no obstacle to wealth, and the ability to feed half the rest of the world, is no guarantee of wealth. I keep pointing this out in the context of people arguing that “we must preserve farmland from urban sprawl”. This is certainly a factor in PREVENTING a thriving, specialised urban economy from emerging. (Refer McKinsey 1998, “Driving Productivity and Growth in the UK Economy” – they say “Silicon Valley” could never happen in an economy with strict controls on urban growth).

    Maybe the world will develop and “populate” one day to a point where farmland will generate high enough “rents” (ECONOMIC rents) that it would lower its value to convert it to urban land. In this case, being a food exporter MIGHT make a nation “wealthy”. But we are NOWHERE NEAR that point today. The “terms of trade” have gone against the food exporters and in favour of the manufacturers, for 6 decades now.

  3. Any engineer worth his salt will tell you that complexity kills. A complex system has a lot of things to go wrong, is difficult (read expensive) to design and sometimes impossible to test before deployment. That is why we engineers learned to value simplicity
    Striving for complexity for the sake of it is just wrong.
    With complexity comes dependence – whether we like it or not. True, when everybody is specialised we are all better off – provided that those who supply us with various vital necessities are wiling to trade with us. When these links break down the whole system becomes a terrible mess.

    • “A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away.”

      Antoine de Saint-Exupery

    • vk and Nogen,

      I’m not sure if you’ve both understood the key difference between complexity (let’s also use the terms diversity) in a network system, and the complexity you describe in common products.

      For example, a simple electrical network would hove one generation plant, one transmission line and one customer. But if the any of the production infrastructure fails, there is 100% loss of power.

      In a complex electricity distribution system there will be multiple generators, multiple transmission lines going to, and between, multiple customers in a fully linked ‘complex’ network. If any of the key links in the network fails, there is likely to be other routes for power through the network. Complexity in networks creates a type of built-in ‘redundancy’. So rather than there being lots to go wrong (which I guess is true in proportion to the size of the network), there are many alternative options for the network to function when portion of it fail.

      More here.

      • Rumplestatskin,

        My point was “Striving for complexity for the sake of it is just wrong” and I still stand by it.

        We obviously learned to use tools more complicated than a sharpened sticks. We demand certain features and we are prepared to deal with the complexity that comes along with these features. Yes, redundancy is a design feature and yes, we are willing to pay more for redundancy.

        Having said that we can estimate in advance what are the costs of a redundant electricity network and factor that in when we make the decision to build it. Our knowledge of the electrical an mechanical systems allows us to make these estimates reasonably well.

        Now enter the world of CDS, CDOs and other fancy abbreviations from the “financial engineering” (hate that term) world. Whoever designed them probably had no idea what will happen when they all blow up. Whoever ought to regulate them certainly did not foresee the mess we are in. Creating complexity in areas not fully understood (i.e. Economics) is a recipe for disaster. This is the sort of complexity we can live without.

  4. So, it turns out again that education and science is the key for economic progress as there is no knowledge today without good education and research. Then the most important question is whether the education and research has to be run as a private business. The answer is already given by Joseph Stiglitz long time ago and it is negative, because of the social collective character of knowledge and the huge external benefits from public education and research. I just wonder why we don’t accept the fact that if knowledge is a collective factor for growth then part of the benefits from the economic growth (profits) should be shared with the whole society through taxes and spend mainly for good public education and research? Obviously the politicians in USA and here don’t understand what is knowledge and what is its impact on the growth, otherwise I don’t understand why the governments are cutting the expenditure for the education and pushing it to become a business.

    By the way, the first who defined the science and knowledge as one of the most important productive forces for the economy and its growth was Marx. Harvard researchers are very much behind his ideas in this matter.

    Here are some quotes about the progressive impact of the scientific knowledge on economy:

    “In the same way as the production process becomes an application of scientific knowledge, so, conversely, does science become a factor, a function so to speak, of the production process. Every invention becomes the basis of new inventions or new, improved methods of production. It is the capitalist mode of production which first puts the natural sciences to the service of the direct production process, while, conversely, the development of production provides the means for the theoretical subjugation of nature. It becomes the task of science to be a means for the production of wealth; a means of enrichment.”

    “With capitalist production, therefore, the scientific factor is for the first time consciously developed, applied, and called into existence on a scale which earlier epochs could not have imagined. ….The development of science alone — i.e. the most solid form of wealth, both its product and its producer — was sufficient to dissolve these communities.”(Economic Works, 1861-1864, Marx)

  5. In the developing countries complexity is imported if there are:
    1. low wages
    2. a sufficiently educated workforce and 3. political stability
    They then get the latest technology based manufacture as they get the new factories and equipment.

    Their time will pass as their economies increase in average age of plant as they must over time and growth and as wages increase so land, factories and labour cost more.

    The patent holders control the ability to export product into developed countries and their role must also be considered. They take the maximum level of profits they can consistent with the cost of land, equipment and labour. They face problems in enforcing their patents against domestic rip-offs but much less so in relation to imports into developed countries.

    The question for Australia is:
    “Where is the strategic plan for 50 years time when the Pilbara runs out of iron ore and our mineral export income washes off?”

    Compound growth in mineral exports is not infinite. Just pull out a chess board and a kilo of rice, put 1 grain on the first square, 2 on the second and keep doubling for each other square to see the eventual outcome.

    • The answer is always good education and research. There is no other way of sustainable economic growth after mining. Germany and Switzerland have some of the best tech universities, that is why they can keep exporting when China became the global manufacturer. The long term benefits for growth from the good education and research are indisputable, but the incentives today are for short term policies and fixings. Watch what is going at the universities today to see what to expect for the future.

      • “…..the best TECH universities…..”

        Important point.

        Starry-eyed young idealists leaving high school and going straight to uni to do some touchy-feely arty-farty trendy P.C. degree, are worse than useless, a DRAG on the economy and society.

        I have seen some truly blood-freezing statistics about the kind of degrees that young people have been choosing for the last few years, when there are important industries crying out for engineers and scientists and so on.

        Most of our kids will be coming out of school believing that “science” is just something you do to be able to provide evidence against “global capitalism”.

        • Phil

          crying out for for engineers? I’d say only a narrow focus in a narrow area is. The rest seems to ignore us (engineers). I mean really, how many people are involved in writing the bios of HD controllers anymore and how much is a thin tip supporting a massive inverted pyramid nowdays?

          mass production and distribution means we need less engineers … except say in mining

      • That’s a bit rough Phil, we can’t all be engineers and rocket surgeons, but I do see your point when every element of education has been “Academised” i.e what was a Diploma level course is now a multi-thousand dollar Bachelor’s degree.

        There is an intangible value to the arts – I would (and intend to in my retirement) support scholarships for music conservatories, painting/sculpture Arts programs, as well as social/anthropological studies.

        I studied history and politics before economics and finance and believe myself better off for the experience to place the latter skills in context (as opposed to studying theology, I mean, economics by itself).

        But before all that, yes, Australia (and NZ?) could well do with a shift in focus to the science and tech degrees (and going by my and other’s experience, an increase in efficiency on how this education is delivered)

        • I’ve seen it described as ‘qualification inflation’, or something like that.

          Another reason for a shift towards ‘softer’ degrees could be because they are cheaper to teach than full blown technical subjects, which require all sorts of expensive equipment, materials, support staff etc.

        • If the upper class twits who actually made Britain into the world’s greatest empire, had received the kind of “Arts” education that universities are giving the young today instead of the CLASSICAL education they DID get, Britain would have been beaten to death by Scotland and Wales, let alone the French and the Spaniards.

          I understand there is a small institution somewhere in Australia called “Campion College” that MIGHT still be WORTH attending for an “Arts” education.

          I am BTW, an avid reader of “Quadrant”.

        • The government should do more to incentivise STEM education, maybe by kicking in a bit more for commonwealth supported places in those faculties.

          However, all the incentivisation in the world won’t make people smarter, but it may just turn the intelligent high school graduates away from the dubiously named “Business” and “Commerce” degrees.

    • an excellent article and I feel I’ve read it before. The nature of which is what drove me to bring my small steam engine in to work to show our educated lot some simple basics. None of the highly educated group I work in had seen one or even understood how it worked.

      Then there was the discussion about would the metho burner set off the fire alarms (metho burns smokelessly) while noone questioned the toaster (which has set off the alarms)

      I work in the area of Arts Education and Law … isn’t education great

  6. Alex, here is where the invisible hand has brought us. Did you see it and do you know it? Watch it, it is very important to know it. We are living in 21 century, not in 19 or 20 and the reality unfortunately is not what is written in some classic essays.,_Inc.

    • Lori, I’m afraid I don’t see a propaganda film as being more “real” than an essay that describes in meticulous detail the complex series of processes that go into making a pencil.

      I don’t doubt, BTW, that there are some nasty practices involved in agribusiness, and there are probably a lot of things happening that the more squeamish might find distasteful, even if they are humane. But I don’t see why you think this aspect of reality has a greater claim on our attention than any other. If it disturbs or angers you, fine, go ahead and do something about it. Lobby the representatives of whatever legislature there is in your country to implement or reinforce legislation to prevent ill treatment of animals. That would be a good thing to do.

      But really I don’t see how the maltreatment of animals in particular farming practices has any relevance to the topic of this thread, which is the advantages accruing to society from specialization.

      • There was an excellent letter to a recent “Quadrant” Magazine, asking how vegans reconcile the deaths of insects, snails, lizards, frogs, fledgling birds, rodents, etc etc as fields are prepared, ploughed, and harvested of their life giving crops.

        • Yes I’ve often wondered about the hypocrisy of that position, plus the paradox that a meat eating culture is more environmentally stabilising (not perfect in the short term, but systemically speaking) in terms of resource use than a vegetarian one (i.e you can only sustain a smaller omnivore population).

  7. I think that (because I also happen to like the Urban Metabolism framework) that a nation should be compared to an eco system. Robust Eco-systems are filled with complex things. The more simple an ecosystem is the more sensitive it is to collapse and unable to regenerate.

    Like my old favourite author suggests “specialisation is for insects”