Of all the tasks given to national statistical agencies, measuring
productivity is one of the most difficult. There is an immense, intricate
infrastructure of data definition, collection, and management required to
produce estimates of annual growth that typically are around one
or two percent a year, plus or minus.
As countries like Australia move from an industrial economy to an
economy that is largely made up of various services, the challenges in
measuring and understanding productivity growth become more complex. And, it
must be said, more interesting. It is, at present, much harder to capture
output and economic activity data in the digital economy, so statistical
agencies are developing new data sources that tap into sources like business
tax information, accounting software and other digital data.
This may be one reason for the productivity
slowdown we are currently in, a topic that is currently the focus of much
attention that will be in another post. If output is not being recorded or is underestimated
while the workforce is increasing, the growth rate and level of productivity
will also be underestimated.
In an agricultural economy, the two resources that determine growth
are land and labour, with a rising population driving increases in output. If
output and labour input move together there is little or no increase in
productivity, because productivity is the ratio of inputs and output.
The creation of an industrial economy required lots of investment in
the built environment, infrastructure, and machinery and equipment. The stock
of physical capital from that investment, when combined with labour, resulted
in a higher rate of productivity growth than previously possible. As that
higher growth rate was sustained over decades of industrialization it allowed
incomes and living standards to rise.
The benefits from increasing productivity are similar to the gains
from compound interest, small annual amounts can, over enough time, lead to
large gains. However, because it is incremental and compounds slowly,
productivity is not really suitable as a policy target. Unlike other targets such
as business investment or employment, the long-term nature of productivity
growth means year-to-year rates should not be used as a policy target. There are
also short-term fluctuations due to the business cycle, as the productivity
growth rate tends to move up and down with GDP growth rates.
The reason productivity is such a central idea in both economics and
politics is this role it plays in long-term economic growth and development. Productivity
growth comes mainly from invention and innovation, and these in turn come
mainly from new knowledge in the form of scientific and technological
discoveries.
This is why the political agenda linking R&D and innovation to
productivity and incomes and living standards is so attractive, to those who
understand the dynamic and value science. It seems to be a bottomless source of
growth and prosperity. This is, of course, not really the case, because to get
productivity growth requires substantial investment. Productivity is not a magic pudding.
Investment in workforce skills and training, known as human capital,
has joined physical capital as a driver of productivity. Thus education in all
its forms, from preschool to postgraduate, has become an important part of the
policy mix that aims to improve Australia’s productivity performance. In a
knowledge economy intellectual capital is as significant as physical capital.
The graphs below are from a 2016 Austrade publication,
and the international comparison shows Australia is not doing too badly across
the industries that make up the economy.
The high rate of productivity growth in the 1990s is attributed to
the rapid development of information technology and telecommunications (ITC)
industries, both from the surge in investment in these industries and because
these are what is known as general purpose technologies. Their products and
service are widely used in other industries, and are adapted for a wide variety
of uses.
The expectation is that new technologies like nanotechnology,
molecular biology and genetic medicine, machine learning and artificial
intelligence, additative manufacturing (3D printing) and remote sensing will be
the drivers of growth and productivity in the coming decades. Obviously, these
industries need highly skilled people and will require capital on a large
scale, so getting the policy mix right is crucial.
But productivity has many different dimensions. The marvelous thing
about productivity is its malleability, it can be used to support a wide range
of policy options (the magic pudding aspect). We hear of productive cities and
boosting R&D productivity by better linking industry and academic research.
It can support arguments for more infrastructure, tax breaks for business
investment or corporate taxes in general, more spending on education, energy
diversification, enabling start-ups and venture capital, microeconomic reform
such as competition policy and labour markets, increasing connectivity and so
on.
All these policies can support improved productivity, if they are
well designed and implemented. That is a significant caveat. However, many of
the current policy settings were put in place when we had an industrial economy
and are not really suited to the knowledge and services orientated post-industrial
economy of the 21st century. If raising productivity is the goal, many
different policies need to work effectively and reinforce each other. And they
will only work over the medium to long-term.
The other issue is the unequal distribution of gains from
productivity growth. Australia, like the US and UK, has had wages and average
earnings growing well below the rate of growth in productivity since the
financial crisis in 2007-08. That spills over into the wider debate about the increasing
inequality of income distribution, its causes and consequences, the reasons for
the slowdown, and appropriate policy responses to the slowdown in productivity.
