With structural changes in production coupled with technological progresses, over time there has been a significant shift in modes of production as well as in patterns of employment. Such a changed pattern is expected to have consequences on employment, resulting in significant changes in task composition of jobs. In the face of Covid-19, on one hand such changed task composition might pose additional challenges to certain categories of workers and on the other, Covid-19 itself is argued to have altered modes of production along with patterns of employment while resulting in new-normal activities, depletion of capital base/savings, changes in occupations, reverse migration, return of international migrants etc.
While utilising different rounds of labour force survey data (2005/06; 2010; 2016/17) of Bangladesh and combining it with US occupation network data (O*NET) along with its Bangladesh-specific adjusted database, it can be inferred that, there has been a gradual decline of routine intensive tasks- tasks involving repetitive and routine activities with lesser involvement of analytical skills. There has been a significant decline in earnings of routine tasks as well, indicating greater returns towards more skilled and lesser routine intensive activities.
An in-depth regression based analysis of education premium reflects a steep increase in earnings premium for those with tertiary education with the premium being the lowest for those with primary education only. In case of skill level, though we observe an increase in high skilled workers (3.6 per cent increase from 2005 to 2016), majority of the work force still consists of those engaged in mid skilled occupations (47.5 per cent) with only around 8.9 per cent being in high skill occupations. In particular, on one hand only 1.9% workers are found to be in managerial position whereas on the other, a significant proportion are engaged in craft and related trades (20 per cent), service and sales work (17.2 per cent) with another 43.6 per cent in low skilled occupations like those of elementary occupation (20 per cent) and skilled agriculture (23.6 per cent). This overwhelmingly high concentration in mid and low skilled occupations, most of which are without formal contractual agreements and job security is argued to have made the workers particularly susceptible to any sudden shocks like that of Covid-19. In addition, occupations which involve lesser routine intensity in terms of task composition have been experiencing a decline in returns and these groups could be the worst sufferers. Research has revealed that, Covid-19 like shocks could have serious implications on the income level of those of low and mid skill and for occupations involving more routine intensive tasks and might lead to, for example as high as 23 per cent of non-poor engaged in elementary occupation, 22 per cent in skilled agriculture, 13.4 per cent of those in craft and related trades, 15.4 per cent in plant and machine operators to fall below poverty line. A similar degree of vulnerability can be observed for daily labours in particular and Covid-19 like shock can put as high as 24 per cent of otherwise non-poor daily labours into poverty. From a different perspective, a decomposition analysis (Shapley Decomposition) reflects that earnings inequality in Bangladesh is driven primarily by 'between group' factors where institutional issues (i.e. 'within group' factors) played a key role - thus in the face of this pandemic, in the absence of trade unionism and minimum wage legislation in many of the sectors, the loss in earnings and job loss can be quite substantial.
Besides, increased importance of touch-less transactions due to this pandemic might result in quicker adoption of 4IR related technologies, leaving particularly the youth labour force with little or no time to prepare.
Though not conclusive, one important long term implications of this pandemic could be on skill formation as the new-poor households might be unable to invest in education of children/youths and some of them might even drop out of the education system. In this context, our RIF (re-centered influence function) Decomposition has found increased importance of task content in explaining earnings inequality-therefore adverse effect on skill formation might eventually translate into increased earnings inequality.
Against the backdrop of pre-domination of low and mid skilled workers mostly with primary and secondary education, the negative consequences of this pandemic on the labour market of Bangladesh can be far reaching. The short term solution though lies on stimulating private investment and efficient implementation of incentive packages of the government to the SMEs in particular, in the long term focus should be on skill formation and for effective utilisation of such skills in the labour market.
However, it is not only low skill content, presence of skill mismatch is also another challenge of the labour market of Bangladesh, which needs to be addressed through effective policy interventions. Despite of relatively moderate increase in earnings of mid skilled jobs, employment in those has increased significantly, which can be thought of as further evidence of skill mismatch. With low level of private investment and job creation due to Covid-19, in order to deal with such skill mismatch policy focus should be given to align education policies with skills demanded in the labour market.
In order to deal with the dual challenges of 4IR and Covid-19, and with declining importance of routine intensive tasks, greater emphasis is needed towards skill biased training programmes, particularly those involving cognitive skill. Furthermore, given the inequality inducing effect of this pandemic, with pro-poor effect of routine task intensity, training involving more analytical and interpersonal skills would be equality inducing so the policy focus should be directed towards that end.
Sayema Haque Bidisha is Professor, Department of Economics, University of Dhaka. She is also Research Director, South Asian Network on Economic Modeling (SANEM) email@example.com