Growth in Canadian gross domestic product (GDP) and labour productivity have been declining since the 1960s, according to Statistics Canada. We trail the developed world, StatsCan says – in fact, the numbers say we’re no more productive today than we were in 1971.
How is this possible when computers, the digital revolution and network connectivity have revolutionized the workplace in Canada? Does anyone seriously believe that Canadian workers today are no more efficient (i.e. able to generate more economic output per hour) than our grandfathers?
Perhaps the problem lies not with Canadians but in the way economists measure productivity.
The standard productivity model employed by Statistics Canada is derived from work American economist Robert Solow did 60 years ago. Solow assumed, for analytical purposes, that the national economy is a closed system – his productivity model doesn’t include international trade. The model also assumes that productivity involves the creation of economic goods and services in factory-type production, where there are recordable inputs (capital and labour), an identifiable engine of production and measurable outputs.
At the macroeconomic level, national output is the sum of the value-added outputs created (avoiding the double accounting of what economists call intermediate goods, or the products created to be used in the creation of final products).
Productivity statistics are used to monitor internal economic performance and to compare performance between nations. The most recognizable performance metric for nations is GDP.
But even Solow began to have doubts about his model. “We can see the computer age everywhere but in the productivity statistics,” he admitted in 1987.
Why did he have doubts?
The answer requires us to accept that modern post-industrial economies are in the midst of a paradigm shift.
Production is no longer confined to centralized factories employing physical capital (tangible assets). Intangible forms of capital, particularly software and digital assets, are decentralizing production while creating a productive machine that doesn’t rely on increased mechanization. Rather, it employs large amounts of intangible inputs and outputs that are difficult to measure and are often exchanged in the Internet, beyond the scrutiny and jurisdiction of governments.
These anomalies are referred to as Solow’s Residuals — “the residual growth rate of output not explained by the growth in (measurable, tangible) inputs.” Solow’s Residual is widely assumed to be an indirect measure of technological progress. Economist Moses Abramovitz has described it as a “measure of our ignorance.”
The productivity problem is rooted in an industrial bias found in official government and statistical circles. Until recently, all productive growth was assumed to involve physical capital. Intangibles, on the other hand, were misidentified as intermediate goods. That meant the value of intangibles (which now dominate our economy) was excluded from the conventional outputs, leading to a significant underestimation of output and productivity in most developed economies.
Economist Paul Romer acknowledged the growing importance of knowledge and technology in the economy more than 25 years ago. He demonstrated that these factors improved productivity efficiencies. An improved productivity analytical framework attempted to measure and accommodate technological change.
In the 1950s, Solow’s classical assumptions were perfectly reasonable. The national economy could be fairly accurately modelled by aggregating industrial output, fixed capital inputs and productivity.
But in an economy dominated by intangible assets and the globalized production of goods, the standard productivity model can’t simply be adjusted on the margins. Productivity analytics needs a total revision in foundational assumptions, starting with a more relevant model of the knowledge economy and reliable intangible investment data.
We need to start at the bottom, with more robust treatment of intangible assets at the firm level by managers and their accounting staff.
Improving productivity is widely believed to be the only way a nation can increase purchasing power and therefore improve the national standard of living. So getting our statistical systems to adapt to the new economy could have far-reaching consequences.
But fixing it means more than fiddling. It probably means reinventing productivity analysis from scratch.
Robert McGarvey is an economic historian and former managing director of Merlin Consulting, a London, U.K.-based consulting firm. Robert’s most recent book is Futuromics: A Guide to Thriving in Capitalism’s Third Wave.