Labour Force Survey 2022: Stats for thought

The Bangladesh Bureau of Statistics recently released some (provisional) thought provoking data highlights from their Labour Force Survey (LFS) 2022. The structure of employment shifted towards agriculture when industrial growth was the fastest. The services sector provided some cushions at a slower rate. More women entered the labour force than men and they found employment in producing their own goods in rural areas. A shift in employment towards the rural economy associated with increased labour productivity divergences between industry, services, and agriculture.

A structural disconnect

Employment has expanded. Employment growth picked up from 0.7% (Compound Annual Growth Rate) between the 2010 and 2017 Labour Force Surveys (LFS) to 3% between the 2017 and the 2022 LFS. The number employed increased by 10.1 million (2017-22), slightly higher than the 10.03 million added to the labour force. The number "openly" unemployed shrank from 2.7 million to 2.63 million. The 3.6% open unemployment rate is not a very meaningful number in informality dominated labour markets where most people can hardly afford to not do anything to support their livelihood.

The structure of employment moved in a direction opposite to the structure of GDP in the last five years. Agriculture has grown as the largest employer at a time when the share of agriculture in total GDP declined from 13.6% in FY17 to 11.7% in FY22. Agriculture's share in total employment increased from 40.6% in 2017 to 45.4% in 2022. The share of industry in GDP increased from 32.7% in FY17 to 35.3% in FY22, but its employment share declined from 20.4% to 17%. A 9% average annual industrial growth associated with 3,50,000 fewer jobs in the sector while 5.1% growth in services associated with nearly 3 million more jobs.

Neither industry nor services growth is associated with urbanisation of jobs. The share of urban employment in total employment declined from 27.9% in 2017 to 25% in 2022 as annual employment growth in urban locations slowed from 4.6% in 2010-17 to 0.8% in 2017-22. This could be related to a decline in employment growth in services (from 2.6% between 2010 and 2017 LFS to 2.4% between 2017 and 2022 LFS) and a decline in the level of employment in industry (between the 2017 and 2022 LFS).

The resurgence of employment in agriculture, domestic labour shedding in industry, and a slide in employment growth in urban locations, when the working age population is increasing and in contrast to what happened in the seven years preceding it, are puzzling. There is some evidence of growing labour-saving automation in industry. Almost all the regulation-compliant garment factories, especially the 157 with Leadership in Energy and Environmental Design certification, have automated major sections of their factories. But it is still too small (between 5 to 10%) to explain a decline in industrial employment and a massive slowdown in employment growth in urban locations.

A more plausible hypothesis appears to be the policy disconnect with the export diversification agenda. Bangladesh has increasingly pursued an industrial policy defying comparative advantage from labour abundance. Domestic market-oriented capital-intensive manufacturing was supported more than unskilled labour-intensive activities, notwithstanding shortage of capital and skilled labour. Made in and for Bangladesh has been the mantra. In exports, an alternative to garments has failed to emerge for over three decades now. Defiant strategy required significant government intervention that obstructed creative destruction as bureaucracy, administrative interferences and rent seeking rigged the functioning of markets.

Female participation – work for pay disconnect

Bangladesh added over 10 million to its domestic labour force during 2017-22, compared with 6.9 million added over 2010-2017. The labour force CAGR spiked from 1.7% during 2010-17 to 2.9% during 2017-22, notwithstanding overseas migration of 4.4 million during 2010-17 and another 3.4 million during 2017-22. The increase in the size of the domestic labour force was driven by a remarkable 5.9 million increase in the participation of women during 2017-2022, compared with only 2.8 million during 2010-2017.

Increase in labour supply in Bangladesh's agriculture was dominated by increased supply of female labour. Women employment in urban locations declined from 4.6 million in 2017 to 4 million in 2022 while increasing in rural locations from 18.7 million to 25 million during the same period. Not only more women grew into the working age but the number of females not in the labour force declined.

The influx of female labour in rural locations was predominantly into work related to own use production of goods. The share of females employed for pay or profit declined from 67% in 2017 to 46.4% in FY22, driven by declines in absolute numbers working for pay or profit in both urban and rural locations.

Female labour intensity in female intensive garments has declined. In 2015, the female to male workers' participation ratio in garments was 64:36, but it declined to 60.8:39.2 in 2018 (CPD 2019) Cross country evidence shows that export-oriented sectors have lower gender discrimination measured in comparable worth. Firms in competitive export markets cannot pass higher costs of hiring just men to consumers. But in Bangladesh firms seemed to have found machines more profitable than female labour to those who allegedly lack the skills to handle advanced technology.

These show improving outcomes for women takes more than removing barriers to female participation in the labour force. Despite substantial progress in closing gender gaps in human endowments, though still riddled with norm-based constraints, there has been a reversal in the type of work women are doing in increasing numbers. Females moving from urban to agricultural settings may be escaping the higher levels of friction in urban areas (e.g. long commuting hours due to congestion). But home production is not as empowering as employment in urban locations in manufacturing and modern services.

Intersectoral productivity divergence

A reversal of the Lewis turning point in Bangladesh is noticeable in the past half a decade. Average labour productivity growth (real value added per worker) declined from 5.8% in 2010-17 to 3.3% in 2017-22. This was primarily driven by a 2.3% (CAGR) decline in labour productivity in agriculture, compared with 2.9% growth during 2010-17. This dwarfed the increased growth in labour productivity in industry from 9.4% in 2010-17 to 9.6% in 2017-22. Labour productivity growth in services declined from 5.6% to 2.7% during the two periods respectively.

Any student of Economics 101 knows that when you observe an expansion in output in a sector with shrinking employment, it constitutes evidence of surplus labour. This means that there were more people working in industry than needed in Bangladesh, contrary to the observation of Arthur Lewis in the 1950s developing countries typically have a surplus of labour in the rural agricultural sector which fuels industrialisation, with workers moving from agriculture to urban areas and industrial jobs.

This happened in Bangladesh prior to 2017 but was reversed in the past half decade! Employment did not graduate to modern services either but moved back to agriculture where average labour productivity declined from 8% of average labour productivity in industry in 2017 to 4.4% in 2022. Average labour productivity in services relative to agriculture rose from 4.9 times in 2010 to 7.6 times in 2022, although declining from 47.3% of average labour productivity in industry in 2017 to 34% in 2022.

The increased divergence in intersectoral productivity could be the result of hysteresis from Covid in 2020 when a large number of urban households migrated back to their rural roots. This may subsequently have been reinforced by the rise in direct and indirect cost of living in urban locations in 2021 and 2022. However, data on industrial real wage and labour productivity growth do not suggest industrialisation in Bangladesh became potentially less profitable. Average labour productivity exceeded real wage growth in industry during 2017-22. Studies have found industrialisation makes labour scarcer. Wages increase leading to a shift away from labour-intensive towards capital-intensive production. While the latter is likely to have happened in Bangladesh, it was not associated with increased real wage growth in industry.

There may be more to it than labour saving technological change in industrial production and domestic market oriented industrial policy. More recently economists have focused on differences in productivity across agricultural and non-agricultural sectors and barriers to structural transformation that include costs of transportation, skill acquisition, and cultural factors. Barriers to labour mobility from the subsistence farming hinterland to the modern sectors would help explain why employment in Bangladesh's subsistence sector remains high, for women in particular, but not why it has increased.

Policy implications

There is clearly a desire on the part of the government to play an outsized role in steering the development of industries and employment. Doing this with bureaucratic structures has so far not demonstrated a high level of competency in identifying successes and backing efficient outcomes.

One policy lesson may be to avoid overemphasising "solutions" tied to deficiencies in infrastructure, energy, human capital, and governance. These solutions help industries by using tariffs and subsidies to compensate so as to buy time for the deficiencies to go away and the industries to become competitive without clarity on how those deficiencies will get fixed and how long the tariffs and subsidies must stay.

A tension exists between promotion of changes that enhance productivity and those that increase the quantity and quality of jobs. Going forward, a more promising target for balancing the two is likely to be services. Putting all eggs in manufacturing-led import substitution and exports basket may have outlived its wisdom.

There is a much bigger borderless global market for services where Bangladesh is building a reputation. According to the Oxford Internet Institute, Bangladesh is now the second-largest supplier of online labour globally, with a 16% share of the global online workforce, following India, which has a 24% share. Labour abundant countries generally are poised to gain dynamic comparative advantage if they invest wisely in building human capital. Bangladesh must not miss out in scaling up the penetration in the global online labour market it has already achieved.

Zahid Hussain is former lead economist of the World Bank Dhaka office