Back in the late 1980s automation was already well advanced: robots could perform many of the functions they perform today, CNC lathes could be programmed offline, and technologies such as the Cadbury Wispa bar high-speed flow wrapping line (that my team at CCL developed) was as much state of the art as anything that has since been developed. What happened in the late 80s was a virtual block on technological development and continued adoption by only a few selected sectors, such as the automobile industry. The biggest change was outsourcing – first to China and then to other Asian countries, Eastern Europe, and Mexico. Cheap labour was seen as less risky and easier to sell within the corporate mindset than technology, so investment in advanced automation slowed down – in some cases to a stop. Manufacturing in general inched along, but still across the USA and much of Europe far too many people were still packing things by hand and undertaking other routine tasks in a largely manual way. They were doing it with faster computers, but the manual processes were largely just being made easier.
The figures largely speak for themselves. Amongst all the talk about labour-displacing technologies (LDT), there is scant attention given to how much economic growth actually depends on the amount of labour in terms of per capita hours worked. If we look at the OECD’s figures for GDP per capita (wealth created per person or GDPP) and compare the situation back in 1985 with the latest available figures (for 2016), there are some surprising outcomes.
Let us take eight countries – Australia, Canada, France, Germany, Japan, Korea, the UK, and USA. We could assume that the importance of information and telecoms (ICT), other investments, and other automation factors would grow over time and the relative importance of total hours worked decline. Technology would thus be the driving force for productivity. However, this is not the case for five of the eight economies – France, Germany, Japan, the UK, and USA. For these countries the essential vehicle for growth has been labour, not technology. This is partly explained when we look at total annual GDP growth levels, which were more than twice as high in 1985 than 2016. In fact, ICT investments in 1985 were higher in all but one country (France) in 1984 than 2016.
The radical change in contributors to output is very recent. Labour costs have been rising for many years in virtually all the favoured production countries – such as China and Eastern Europe. But as they rise in one country, manufacturers and call centres have tended to relocate to countries where labour remains cheap. But this strategy has its natural nemesis. Even in Bangladesh, where wage rates remain cheap, employment conditions are generally improving quickly so that associated labour costs are rising faster than wages. US protectionism is also driving US companies to reshore production. As companies do this then the investment decisions centre more on LDTs than training up shop floor or routine office workers. So, if there is an upward trend in LDT it is recent and, in many cases, very sudden. It will catch many HR professionals unprepared and there is nothing in the repertoire of HR rule book and tactics to cope with it.
Looking ahead to the next 2 years, the biggest impact on jobs will clearly be an economic recession. This time though, those who lose their jobs will probably never re-enter the labour market, unless they can reskill. As liquidity creeps back during the next upward economic swing, investment will be made heavily into ICT and the old world of labour-intensive production will be over, even across much of the services sector. Machines may not do it with a smile, but they normally do it cheaper in the long run, and much better.