The Bangladesh Story in Eight Charts

There is much talk about short-term economic challenges faced by Bangladesh. That is expected. There are indeed several immediate challenges that need to be addressed -- unrelenting inflation, near-stagnant investment, factories closing, and jobs disappearing.

But there are also long-term challenges that we must attend to. We need to do this at the same time as we attend to the more immediate challenges. I am talking about the agenda for transforming the economy. Not a little bit here and a little bit there, but structurally and in a profound way.  

This agenda is long-term and yet urgent. The world is moving fast, and we can’t afford to procrastinate.

In this piece I will tell a story using eight charts. The story is eye-opening. It is also disturbing. The story I tell will confirm the urgency of this agenda.

The first chart shows the place of garment exports in Bangladesh’s export basket. Its presence is overwhelming. Four decades ago, around the time it started its journey, Bangladeshi garment exports accounted for less than 5% of total exports. A decade and a half later, i.e. by the end of the 1990s, this proportion had reached about three-fourths. In the last decade and half, it has hovered between 80-85%.

Chart 1: Garment exports as % of total exports: 1983/84 to 2024/25

This heavy dependence on a single product is a structural weakness. Of course, there is still scope to increase garment exports and there is indeed scope to diversify within the broad category of garments.

But Bangladesh needs to go beyond garments and diversify its exports, especially towards more sophisticated products such as electronics. The dynamic economies of Asia had also started with simple products but pivoted to more sophisticated products much earlier in their development trajectory. Bangladesh is way behind in making this transition.

This brings us to Chart 2 which shows Bangladesh’s comparative performance on the Economic Complexity Index. This important index has been constructed by a group at Harvard University which studied the technical characteristics of different products exported by countries and assigned complexity scores to these products based on such features. Sophisticated products such as machinery, pharmaceuticals or, various electronic products get a high complexity score while simpler products such as garments score lower.

The scores for individual products in a country’s export basket are then aggregated using appropriate weights to come up with a composite score for the country.

Chart 2: Country ranking on Economic Complexity Index

Complexity is important because a country that starts making complex products acquires skills in a wide variety of areas. Businesses do acquire some skills when they make and export simple products such as garments.

Our garment exporters did learn how to deal with demanding foreign buyers, meet their exacting quality standards and delivery deadlines, and respond to fast-evolving fashion trends. Banks learned how to finance international trade and millions of workers learned how to transition from the informal and more relaxed setting of their rural households to the formal routine of factories.

That’s all true. However, when businesses go into more complex products such as electronics and pharmaceuticals, they acquire a much wider set of skills, which can then be used to branch out into other, equally complex products. Entry into a complex product is not just an important step towards diversification. More importantly, it opens the door to more complex products.

Chart 2 tells us that Bangladesh ranks considerably lower on this front than comparators such as India, Vietnam, and Malaysia. In 2020, for example, Bangladesh ranked 93 on this score, significantly lower than Vietnam (52nd), India (46th) and Malaysia (24th).

Interestingly, Bangladesh did better than Vietnam in 1995 when it ranked 89th while Vietnam lagged with a rank of 107. In other words, Vietnam’s export basket was even simpler than Bangladesh’s. But what Vietnam did after that is striking.

One clue to that is provided in Chart 3, which shows long-run trends in electronics exports as a percentage of total exports for Bangladesh and Vietnam.

Chart 3: Electronics Exports: Bangladesh loses the race, but Vietnam forges ahead

In 1990, both countries had a negligible presence in electronics exports, with such exports accounting for less than 1% of total exports. Since then, the share of electronics exports in Bangladesh’s export basket has virtually stagnant -- remaining around 1% -- while Vietnam galloped ahead. By 2018, its electronics exports accounted for a third of its export earnings. No surprise that Vietnam has rapidly improved its rank on the economic complexity score while Bangladesh has fallen further.

Let us now move from exports to the broader enterprise landscape of Bangladesh.

Charts 4 and 5 are based on data from the economic census of 2001, 2013 and 2024. The census report identifies three types of establishments:

a) Permanent establishments

b) Temporary establishments

c) Economic households

Permanent establishments operate from a fixed, physical location and vary widely in their nature, from a neighborhood grocery store to a large steel mill. Unlike temporary setups, these persist over time, though some may close eventually.

Economic households engage in non-agricultural activities such as cottage industries, shops, or street vending within or outside their homes for profit or household use.

These informal activities operate outside regulatory and social security frameworks, typically with low productivity. Generally, permanent establishments are larger and more productive than economic households, employing an average of 3.77 versus 1.25 people in 2024.

The charts show some significant changes in the enterprise landscape of Bangladesh. There is both good news and bad news. Chart 4 gives us the good news: In the last two decades, the number of economic establishments has increased three-fold at a time when total population increased by 27%. As a result, we had 68 economic establishments per thousand people in 2024, compared to 27 in 2001.

Clearly people have become more entrepreneurial. But the average employment in such enterprises has decreased -- from 3.13 employees in 2001 to 2.59 employees in 2024.

Chart 4: Enterprises are proliferating but not growing

The reason for this is revealed in Chart 5. And this is the bad news: the proportion of permanent establishments has fallen over the same period, from 81% to 53%. Economic households now constitute 42%, up from just 10% in 2001.

The enterprise landscape is now increasingly dominated by household enterprises, complemented by self-employment.

Chart 5: Household enterprises are becoming more important 

Chart 6 (which is reproduced from a recent World Bank report, Frontier Firms and Job Creation in Bangladesh) shows a similar trend from another vantage point. The right-hand panel of the chart shows the steady growth of productivity in the manufacturing sector over the period 2010-2019.

Manufacturing productivity is still low compared to most of our comparators, but it is heartening to see it growing. Productivity in the services sector also grew initially but stagnated after 2016.

We would want labor to move towards the sector with higher and growing productivity, i.e. manufacturing. The left-hand panel tells us that the opposite is happening in Bangladesh.

The share of manufacturing in total employment grew very modestly from 2010 to 2015 and then stagnated at about 20%. The share of the services sector grew slowly in the first half of that decade but accelerated from 2016. By 2018, the year it overtook agriculture as the largest employer, its share of total employment had crossed 40%.

Chart 6: Labor is moving towards sectors with lower productivity

Let us go a bit deeper into the manufacturing sector and explore some data from the last two surveys of manufacturing industries (SMI), done in 2012 and 2019. The SMI covers manufacturing enterprises which employ at least ten people. It groups enterprises into four categories: micro (employing between 10-24 employees); small (25-99 employees); medium (100-249 employees) and large (250 or more employees).

Chart 7 (reproduced from the report of the Government’s Task Force on Economic Re-strategizing) summarizes some key data from the surveys. The results are striking!

Chart 7: A hollowing-out of the middle

We can see that, between 2012 and 2019, there has been a significant change in the size structure of Bangladesh’s manufacturing sector. The medium-scale enterprises, i.e. those employing between 100 and 249 employees, have seen a huge drop in both their number and employment -- by almost 50% in both cases.

The opposite is true of small enterprises which have grown in importance both in terms of numbers and as a source of jobs.

In the case of large enterprises, the two variables have moved in opposite directions -- less enterprises but more employment. The average large enterprise has become substantially larger in employment terms. While this could be interpreted as a good trend, the hollowing up of the middle is a matter of concern since the medium scale enterprises are often the source of dynamism in an economy.

The final chart provides a comparative picture of innovative activity in our manufacturing enterprises, based on data from the World Bank’s latest (2022) enterprise survey.

There is data on four different indicators of innovative activity, i.e. introducing a new product or process innovation in the last three years, spending anything on R&D in the last fiscal year, or using technology licensed from foreign companies. The picture is bleak, not only when compared with global averages but also relative to our South Asian neighbors.

Chart 8: Bangladeshi firms score low on innovation

Bangladesh has delivered high growth rates, manufactured millions of jobs, and achieved major social gains, but without the sort of structural transformation that usually accompanies sustained development.

That is the sobering story coming out of the eight charts.

Enterprise numbers have surged, but they are increasingly small, household-based activities. It is a vast economy of micro and small enterprises. These are vibrant, flexible, and ubiquitous. But their productivity is low and growth prospects poor.

Manufacturing productivity is rising, yet employment flows mainly into lower-productivity services. Innovation indicators, from process improvements to R&D, are uniformly low.

We are now in the fifth decade of making and exporting garments, and still heavily dependent on it. The phenomenal growth of garments signifies the triumph of specialization. But its continued dominance tells us we have not diversified enough. Thus, Bangladesh’s economic complexity ranking has slipped behind peers such as Vietnam that began with similarly simple export baskets.

The East Asian dynamos -- Korea, Taiwan, and later Vietnam - had diversified into electronics, machinery, and chemicals much earlier in their growth cycle.

But it was not just a story of diversification. These economies created conditions that helped, and in many cases, forced their enterprises to continuously upgrade their capabilities and increase productivity.  Knowledge spillovers were supported through various means. Firms learned by doing things on their own, but they also learned from others, including foreign investors and machinery suppliers.

Medium-scale enterprises have been an important set of players in these countries.

These are often incubators of managerial learning and technological adoption, and often the fulcrum of structural transformation. In countries such as Korea, Taiwan, and Vietnam, such medium-scale firms were often linked to lead exporters, including foreign invested companies, or to global supply chains. This is how these firms learned to meet the high quality and delivery standards of demanding buyers in global markets.

To do so, they pooled skilled labor and diffused knowledge amongst themselves. In Bangladesh, medium-scale enterprises have receded rather than expanded. This hollowing-out of the middle is a particularly worrisome phenomenon, signaling a weak ladder for firms to scale up.

It is not that Bangladesh’s economic actors have not learned anything or acquired any capabilities.

Many of them have. For example, our garment sector did learn how to meet tight delivery timelines, quality standards, and financing needs.

In pharmaceuticals, some firms are into cutting edge R&D and drug discovery work. A beginning has been made in designing semiconductor chips for cutting-edge clients in Silicon Valley. But these capability gains are limited to some niches and not widespread.

By contrast, in Vietnam and other Asian peers, capabilities multiplied and recombined, opening new pathways.

Diversification was a big factor. Going into sophisticated products, such as electronics or automobiles, forced these economies to continuously learn about production methods, design, logistics, and technology.

As firms mastered one layer of complexity, they used those competencies to enter adjacent sectors, creating a ladder of increasingly sophisticated activities. In Bangladesh, the frontier of capabilities has advanced but only very slowly.

In brief, profound structural transformation, the basis of sustained development, has eluded us. Productivity gains have been slow, and the economy remains quite vulnerable to shocks.

Bangladesh is adding more activities and workers, but not enough complexity, scale, or technological depth. The economy is busier, but not necessarily more capable. It is learning, expanding, and improvising, but still lacks a broad base of complex production and a strong eco-system for firms to scale, innovate, and experiment.

In this piece, I presented eight charts. But they tell basically one story. On the economic front, Bangladesh’s achievements are real. But the transformation is still incomplete.

Syed Akhtar Mahmood is an economist, previously with an international development agency.