Since the 2008 global financial crisis, the non-farm labor productivity levels have been declining. The annual rate of productivity growth since 2010 has been around 0.5% and is below trend. This pattern of slowdown in productivity growth is applicable for all advanced countries for the last decade.
Several explanations have been put forward to explain this phenomenon. Some like Martin Feldstein claim that growth in real GDP and productivity are understated by official statistics as output of some sectors are hard to measure. But the slowdown has not been confined to only those sectors where output is hard to measure. In the age of the internet of things it seems counter-intuitive that google and other internet utilities have not increased productivity.
The other explanation is that since most productivity enhancing technologies are capital goods, capital investment plays a very important role. Since 2008 due to weak demand and thereby low capacity utilization rates, capital investments have not seen an uptick. So even though employment is improving, new capacities are not needed to accommodate the recovery in demand.
Robert Gordon argues that the productivity increase gained by innovations like indoor plumbing, jet engines and electrification were far more than that of the internet and smartphones.
Keeping the above arguments in mind, growth can come from productivity increases due to better technology, effective organization structures, better institutions, resource extraction, newer labor processes, increase in labor force participation and trade.
Service sector activities are labour intensive and on average is associated with little growth in productivity over time. There is requirement of better capital technologies for improvements in productivity. The Baumol effect which is the increase of wages in a sector with low increase in productivity in response to rising wages in another sector experiencing high productivity. It is the productivity increases which sustains long term growth. Wealth is the increase in productivity in the economy.
The correlation between household debt payments as percentage of income and real output per hour from 1980 to 2015 is 0.149 2. It can be inferred that over time debt increases was not linked to long term ability to pay back due to its low association with productivity levels. The correlation between increase in annual median wage and annual labour productivity since 1988 is 0.351 3. So over time there was delinking between productivity and wages, which was not sustainable. The higher wage also enabled them to take on more debt increasing cost of capital for productive investments. As service sector grows in size, labour productivity increases require higher capital investments compared to manufacturing sector.
The arrival of internet and the increase in globalisation has boosted wages in certain service industries. The Baumol effect may have bid up wages in other sectors compared to productivity increases it achieved. At the same time expansionary monetary policies to counter busts and increased borrowing from the world led to low cost of credit. A lower cost of credit led to bubbles and malinivestments. The crowding out of credit by speculative sector led to decreased investments that would lead to long term labour productivity. Big data, artificial intelligence technologies and automation hold the capability to unleash productivity in the service sector.
2 https://fred.stlouisfed.org/series/TDSP, https://fred.stlouisfed.org/series/OPHNFB
3 https://fred.stlouisfed.org/series/TDSP, https://fred.stlouisfed.org/series/MEHOINUSA672N