When we see a storm coming, we do our best to prepare. The next two decades are likely to bring some very big threats for jobs right across the world. For millions this will be equivalent to a storm, flood or drought. Yet many individuals, firms and nations are not preparing. We must prepare not only because we must address the threats but also because these threats are coupled with unprecedented opportunities.

This short blog shares some ideas on how Bangladesh could create a shared, living, national system for responding to these challenges, providing workers, businesses and the nation with accurate understanding of the current facts, critical future trends and guidance on how to respond. It links to work we are doing with the UNDP and a2i, Bangladesh government’s flagship public innovation program, to explore the potential of an adaptive system for jobs and skills in the country.

Some background

Over the next 20 years, Bangladesh — like many other countries — faces some very big challenges to future employment — as the combination of automation, artificial intelligence, fourth industrial revolution and other trends, such as the shift to a circular economy, threaten existing jobs in key industries such as garments and textiles. Some of these challenges may be particularly acute. A recent analysis suggested that 47% of the jobs in Bangladesh may be at risk by 2041 in 5 sectors — including 60% of jobs in garments which currently represents 81% of exports.
As in many countries, these challenges may be hard to respond to. Individuals often lack the knowledge or motivation to reskill. Employers may see little benefit for them in training employees for new jobs with someone else. Within the public sector there may be complex and fragmented responsibilities; lack of a shared perspective on vulnerabilities and opportunities; misalignment of policy; and sometimes the misconception among policy makers, business owners and employees that these challenges are not imminent in our context.

The problems and challenges

Technological change has been transforming jobs for two centuries and more. Jobs constantly appear, disappear and mutate. But the pressures may be accelerating along with the various factors sometimes described as the Fourth Industrial Revolution. These are just some of the factors:

• automation of repetitive manual and non-manual jobs.
• the spread of artificial intelligence in its many forms.
• new combinations of digital technology and physical systems for energy, buildings and transport.
• demographic dividend in many developing countries but the window closing around 2030.
• ageing populations in many developed countries but also increasingly in developing countries.
• climate change (and policies to reduce carbon emissions).
• large-scale forced displacement due to human conflicts, disasters and climate change.
• changing consumer demands.
• changing business models.
• policy influences, e. g. pressures for data localisation in the face of lacking disaggregation.

These can have very uneven impacts by sector, gender and geography. For example, automation in textiles has already reduced the female share of the labour force in Bangladesh from around 80% to just over 50%. Furniture has been influenced both by new manufacturing technologies but also by changing business models leading to much more modular production and self-assembly by consumers.

A key lesson of recent years is that generalisations are misleading. Many of the effects are very detailed and contextual, even if there are also some broader megatrends. For example, textiles production is likely to be influenced by new rules in consumer countries on microfibres, nanoparticles, plastics of all kinds; or requirements for recycling, or changing use of dies.
There have been many forecasts looking 10–20 years into the future. The dozens of studies (World Bank, Oxford, McKinsey, Nesta, UNCTAD, PWC), most focused on rich countries but some covering developing ones use slightly different methods, though most combining expert opinions with machine learning. They are far from perfect and it would be unwise to put too much weight on any specific, quantitative forecast. But there are some fairly common conclusions at least in the 5–10 year horizon that point to some of the actions that are needed.

Most commentators agree that there will be an even higher premium on basic literacy and numeracy in the future — continuing a very long-term trend — but now including some other kinds of literacy such as ability to use the Internet. Most agree that there is a high likelihood of automation of many repetitive manual and non-manual tasks, though with some exceptions where perception, manipulation, and creative and social intelligence, play a role. So, for example, tasks requiring subtle dexterity have often been thought to be relatively resistant to automation. But there are signs of some progress now in automating difficult tasks in fields like embroidery, leather work or machine repair, though less in picking tea and coffee and other commodities.
Right digital skills: Most expect there to be greater demand for digital skills, though this is complex. There is bound to be some more demand for generic awareness of digital technology, understanding of coding, data analysis and how to use the Internet. There is also likely to be growing demand for some very job-specific digital skills in fields like AI, VR, AR and analytics. But many digital skills could be in lower demand — as they are superceded by next generation technology, and often these are the easiest tasks to automate using AI.
Skills that are hard to automate: In most countries it is expected that jobs demand will rise in fields such as care, education, tourism, partly because of what economists call ‘positive elasticities of demand’ (people pay more as a share of income for these things as incomes rise) and partly because these are hard to automate.

Shift from credentials to future skills for the job market: Finally, there are some common views on the implications for education. Promoting cultures of learning and curiosity may come to matter as much as specific skills, since the latter may well become obsolete fast. Moreover, attention may increasingly shift from credentials to skills — many countries have a big disparity between incentives around credentials and actual labour market demands, with significant negative effects on graduates in particular whose credentials turn out to be of little use in securing good jobs.

How to respond — possible building blocks for a new adaptive national system

These points are widely accepted. But few countries are yet acting in response to them, for example in their schooling systems. Why is it so difficult to respond? We’ve already suggested some of the barriers, from lack of awareness the challenge of motivation given that gaining new skills usually involves costs in terms of time and provision, inconvenience and uncertain returns. Fixing these problems isn’t going to be easy.

But although we cannot be certain about the details of future trends, we can be certain that governments and organisations need to more prepared to adapt and respond. Indeed, this may become a key differentiator for countries.

Countries that want their citizens to thrive will have to focus on both the supply of labour and the structures of demand. On the demand side attention to job creation, and the quality of jobs demand needs to be part of the answer, and connect to broader economic strategies to boost investment and productivity. Past orthodoxies — such as the role of manufacturing in driving development, taking in unskilled workers, growing capacity and export earnings — may be becoming obsolete. On the supply side, key roles will be played by all the many organisations involved in training and education, including families and employers.

To guide both, Bangladesh will need to mobilise many sources of intelligence to help its labour markets adapt quickly and efficiently. These include data, tacit knowledge, business insights, evidence and experience from other countries. The aim should be that everyone can make decisions based on the best possible insights.

There could be at least five main building blocks for such a system:

• REAL TIME DATA Comprehensive and curated real-time data on current patterns — the state of jobs demand, what skills are being looked for in jobs, pay levels, ideally to be easily analysed by geography and sector.
• FORECASTS assessments of job vulnerability at individual, job levels, drawing on analysis that breaks jobs down into bundles of competences. These become most useful if they can be shared in easily accessible ways with workers, students, schoolchildren, parents and educators.
• OPPORTUNITIES Comparable analysis of emerging opportunities, including growing jobs and potential fields for new jobs — so as to feed into the design of training and other systems, as well as business strategies.
• EXPERIMENTATION AND ADAPTATION Sector strategies for adaptation, involving the key industry bodies, and combining support for adoption of new technologies, new business methods and models, and new skills. To arrive at the right strategies, a fair bit of experimentation is necessary and must be encouraged.
• POLICY National policy strategies involving key bodies such as Ministries, NSDA, universities, polytechnics, schools, companies, industry associations and civil society organizations, to ensure that the powers of the state are aligned with future needs.

If these can be pulled together in a coherent way through meaningful and sustainable public-private-academia partnerships, they should enable action at individual, firm, sectoral and national levels — with exemplars, leadership and a sense of agency for all who face a shifting landscape of threat and opportunity. The focus is not just on planning together but acting together.
Many countries are grappling with similar issues, sometimes helped by global bodies like the ILO, WEF, OECD and the recently announced Generation Unlimited platform of the UN globally facilitated by UNICEF. Commercial firms are expanding which data is available and how it is used — such as Burning Glass, Faethm, Linkedin and others. Policy innovation is happening around mid-career training, new kinds of personal training account and transitional income, with some countries such as Singapore putting in place comprehensive programmes. However, the use of such platforms is limited in developing countries and even more limited for blue collar workers.

But no country has yet created a really effective shared intelligence — which is a necessary if not sufficient condition for navigating the possible storms ahead. This is where attention needs to turn now, pooling knowledge to help people and firms cope with the rough edge of change. Bangladesh is taking its first bold step to explore options for making this kind of adaptive national system into reality through a collaboration of various government stakeholders, industry associations, and development partners. In this regard, it has already started experimenting with about 50 small projects that will inform how the adaptive national system will emerge and mature.

This experimentation and exploration should be interesting for a lot of countries in the same boat. We are looking forward to being part of this exciting journey with the a2i team starting this March in Dhaka.