'Factories won't be fully automated for 10 to 15 years'

The cost of new technology is falling. For example, a logarithmic scale – a nonlinear scale used when analysing a large range of quantities – which used to cost $10,000 some 17 years ago, now costs less than $10. This fall in the price of digital tools suggests that it is likely to change the economy both nationally and globally.

During his recent public lecture in Dhaka, Stefan Dercon predicted that on-shoring value chains – a practice of transferring business operations that had once moved overseas back to the country of origin – in the US, Japan and Europe, will impact income distribution and wage growth in Bangladesh. 

In an exclusive interview, Dercon, also an academic director of the Pathways for Prosperity Commission on Technology and Inclusive Development at Blavatnik School of Government at Oxford, shared his thoughts with The Business Standard.

TBS: Will automation and artificial intelligence put Bangladesh's traditional ready-made garment (RMG) industries at risk?

Dercon: Ever since the industrial revolution began in England, the RMG industries, globally, have gradually been automating. The question is, will automation happen faster given that the technology or kind of tasks involved – those which can be automated – are still performed by low-cost workers who require some agility and experience? It will take a long time before robot arms can actually start doing these tasks. Therefore, by no means is it profitable for factories anywhere in the world to replace labour at the moment. Over time, the RMG industry will get more and more automated and labour intensity will go down. But it will not happen immediately – it will take around 10 to 15 years for factories to become fully automated. Automation will happen gradually and fewer jobs will be created; at the same time, investors should keep in mind that automation will take time to become a profitable venture. Automation will cut a number of jobs and I am worried about the low scale employment in the RMG sector of Bangladesh in the next generation. 

TBS: Recently we met a former expatriate who upgraded his handloom to a power-loom garment factory in his village. He is happy with his income. How do you see these kinds of entrepreneurship?

Dercon: This person clearly found a way of making profit by doing what he enjoys doing. In fact, the economy is full of people who do certain things in line with their aspirations and expectations. Unfortunately, in Bangladesh, there are many people for whom there are not enough job opportunities and very few choices that coincide with their aspirations. However, I actually remain very optimistic that even if a sector like RMG gets fewer jobs, jobs will be created in an economy with the right framework, correct set of incentives and proper focus on being open for new ways of doing things. 

TBS: You talk about diversification of employment. How can modern technology influence agricultural marketing – an alternative source of employment in Bangladesh?

Dercon: I think we should not focus too much on the impact of digital technology by automation. The big impact is more likely to be much better information for farmers to improve their practices for higher yields. Farmers should know about the right use of seeds, fertiliser and water. In addition, access to the right information can reduce wastage and boost productivity. At the moment, most advice to the farmers is quite standardised. Big data and artificial intelligence allow one to get customised advice as per one's needs. This can help farmers with every type of crop. So, it [the benefit] is not mechanisation or automation [solely] but actually things like knowing exactly when to harvest and having the information that a harvest might bring a good price on the market that day. 

TBS: Do you think that ownership of technology can create a monopoly and will affect a society like Bangladesh where the gap between rich and poor is high?

Dercon: It is indeed one of the big risks. By the nature of rapidly reducing costs by having more data and an economy of scale for using data, monopolies are encouraged – or at least a very large market [is created]. If monopolies then capture all the profit from it, inequality is further deepened. To avoid this, keeping pace of digital technology and the emergence of firms, it is important that there is a sensible competition policy, a good understanding – by leadership in the government – of the issues involved and reasonable regulation that keeps a level playing field and eases the entrance of new players into it.

For example, monopoly power over data is not allowed and certain types of data should always be available to all players. For example, rainfall data cannot be sold and one firm cannot control all weather data.