Public ideas
On the scientific method
Because science is under attack from many quarters, many have come to the defence of the scientific method. That is the process of developing a hypothesis, observing behaviour, and devising experiments to confirm or reject the hypothesis.
Joe Walker has drawn attention to academic research which effectively applies the scientific method to the scientific method itself. That sounds a bit self-referential, but it refers to research by Alexander Krauss, of the London School of Economics, that examines the processes that have led to major advances in scientific knowledge. He asks to what extent was the established scientific method applied in these developments.
The results of this study are published in Nexus, the journal of the UK National Academy of Sciences: Redefining the scientific method: as the use of sophisticated scientific methods that extend our mind. To quote from the abstract:
This study reveals that 25 per cent of all discoveries since 1900 did not apply the common scientific method (all three features)—with 6 per cent of discoveries using no observation, 23 per cent using no experimentation, and 17 per cent not testing a hypothesis. Empirical evidence thus challenges the common view of the scientific method. Adhering to it as a guiding principle would constrain us in developing many new scientific ideas and breakthroughs.
Importantly this is not a rejection of the scientific method, nor is it a rejection of the idea of scientific scepticism. But it is a reminder that there is no one path to knowledge.
Is artificial intelligence over-hyped?
In most “developed” countries, including Australia, productivity has been falling over the last 25 years.
One explanation is that we are now in the phase of diminishing returns from the information and communications revolution, but artificial intelligence will release a wave of innovation that will see productivity pick up again.
That reasoning aligns with the idea that economic progress proceeds not in a straight line, but in waves resulting from specific technological innovations – steam, electricity, telephones and so on. These are known as “Kondratiev waves”, after the Soviet economist Nikolai Kondratiev (who Stalin had executed for being cleverer than the apparatchiks in Gosplan).
Writing in Reuters, Felix Martin suggests that AI may fail to boost productivity to any significant extent.
His general point is that while AI is very good at recognizing patterns and predicting trends, it is not capable of explaining the mechanisms that drive system behaviours.
He acknowledges that AI is well suited to some of the slog work associated with administrative tasks, such as searching through legal documents, but the resulting gains in productivity, he believes, will be minor.
He warns that while corporations that take on AI may enjoy tremendous benefits, these do not translate to economy-wide gains. AI may be able to help a firm get a marketing edge over its competitors, but such competition is generally zero-sum. In fact it could even allow firms to gain market dominance, incurring on economies the deadweight loss of industry concentration and monopolization.
Human capital
Early economists, including Adam Smith and David Ricardo, developed three categories of “factors of production”, these being land, labour and capital. Land is a shorthand for all naturally-occurring goods – minerals, water, the atmosphere, plants and animals. Labour includes the employed and the unemployed – everyone from the Deliveroo worker through to the CEO of the Commonwealth Bank. Capital is everything that can produce or distribute goods – machines, ships, stock of finished goods, buildings and computers.
It’s never been a satisfactory categorization, in that it tends to force policymakers to consider each category separately, rather than considering the economy as a complex interactive system. Nevertheless this categorization has shaped the demarcations around university faculties, statistical agencies and government departments.
The most troublesome of the three is capital.
Barry Jones says that we carry in our mind the dated idea that capital is “stuff that hurts when you drop it on your toes”, when so much of our productive capacity is based on skills, knowledge and experience.
For a time some economists added a fourth category, entrepreneurship. This was a right-wing assertion that our prosperity is dependent on the unique contribution of, Übermenschen, worthy of some special return, but for the most part economists refer simply to “human capital” as a subset of capital. The skills, knowledge and experience of the truck driver are no different in kind from the skills, knowledge and experience of the chief executive or the venture capitalist.
Last month the Reserve Bank brought together a number of economists in a conference on human capital. The program, including links to most presenters’ contributions, is on the RBA website.
Most sessions are about what in another context may be called “labour economics” – all useful for policymakers. For example there is an informative paper on graduate outcomes by Lisa Bolton of the Social Research Centre, confirming that that there are significant returns from postgraduate studies. A paper on occupation and industry mobility by Benjamin Mitra-Kahn of the Productivity Commission dispenses with common assumptions about labour mobility: in fact when we change jobs we are quite likely to move to a quite different industry and in a new occupation.
It’s the conference initial paper by Jeff Borland, University of Melbourne and Michael Smedes of the Australian Bureau of Statistics that goes into the development of a specific theory on human capital: ”A review of how economists’ thinking about human capital has evolved and the current state of research on human capital in Australia”.
It discusses the cases for and against including human capital in national accounts (it’s too hard). It covers the way the economic contributions of education, skills and experience, to the extent that they are separable, can be dealt with in human capital theory. And importantly it goes into the contribution of what goes on in the years before children start school.