The World Bank's recent Change of Fabric report provides an empirical account of the sources of Bangladesh's growth acceleration in the past four decades. Bangladesh implemented a series of reforms during 1990-2004 leading to growth persistence subsequently. The report cautions about the historical tendency of growth to slow as income levels rise unless renewed by growth propelling policy and institutional innovations. As reforms lost steam in Bangladesh, the echo from the past began to fade. A "change of fabric" can reset the virtuous cycle of growth.
Growth unleashed by structural reforms
GDP growth accelerated over time, averaging 3.8% in the 1980s, 4.7% in the 1990s, 5.6% in the 2000s, and 6.4% in the 2010s. Manufacturing grew faster than others throughout. The WB report decomposes actual growth into past structural improvements (persistence), contemporary structural improvements, and unexplained (residuals) components based on a cross country model of growth. The model identifies the growth contribution of 12 policy correlates that associate with 87% to 95% of the observed variations in growth across 135 countries during 1970 to 2019. The correlates cover global integration, infrastructure, finance, economic management, and crises.
Bangladesh's growth experience fits these correlates reasonably well until the 2010s. Application of the estimated growth equation to data on Bangladesh shows the typical correlates provided strong impetus between the early 1990s and mid-2000s (Figure 1). During this period, "few countries improved growth drivers as substantially as Bangladesh did", observes the WB report. Fast improvements in infrastructure, greater openness in finance and trade, and stability were salient during this remarkable reform episode (Figure 2). These founded conditions for growth persistence subsequently.
Persistence is no mystery. Reforms led to periods of rising savings, investment, trade, and productivity. Investments increased from 17% of GDP in FY90 to 32% in FY19, supported by increased national savings from 18.3% to 31.1% during the same period. Trade grew 11% annually on average over the two decades preceding the pandemic. Value added per worker increased in all sectors. Agricultural employment declined as a share of total employment and in absolute numbers. Many moved to manufacturing. A much larger number moved to the service sector as the share of urban population rose from 23.6% in 2000 to 37.4% in 2019. Shrinking poverty levels extended growth persistence.
Transformative reforms have been missing since 2004. Stop-and-go policies and procrastination on the grounds that "now is not the time" stalled progress. Yet growth kept rising because of macroeconomic and institutional stability allowing the long-term benefits from previous structural improvements to run their course. The inability of institutions to adapt to a changing economy, locally and globally, through a second generation of reforms made the contribution of persistence to per capita GDP growth decline from 3.5 percentage points during 2005-09 to 1.5 percentage points in 2015-19.
Growth accelerated in Bangladesh while declining in the structural peers. The acceleration over 2015-19 appears to have come from an unidentified source. As much as 3.7 out of the 5.4 percentage points of average per capita GDP growth in 2015-19 are unexplained after accounting for the diminished contribution from policy innovations and persistence relative to the previous decade. The extent of Bangladesh's catch up with peers such as Indonesia, India, Vietnam, Thailand, and China loses much of its shine without this known unknown source of growth. The large residual during 2015-19 deserves unpacking so that mismeasurement can be ruled out in rating Bangladesh an outlier in the history of growth.
Persistence rarely persists
Steady growth is not an autopilot. Few economies remain among global growth leaders over longer periods without recharging. Summers and Pritchett observed in 2013 that "regression to the mean is the single most robust finding of the growth literature". Growth tends to revert to the average through episodes of accelerations and reversals over time. "This in a decade or so is the rule, not the exception across all decades."
The fundamentals of growth can spawn under a variety of institutional environments. Growth in China and India, as in Bangladesh, persisted despite low quality of institutions with a large divergence between the de jure laws and regulations and the de facto arrangements for specific enterprises. The de facto arrangements (closed ordered deals) allow high and secure profitability without the neutral enforcement of the rule of law. The environment specific to the privileged firms is in fact often better than the existing de jure regulatory environment.
Growth in such circumstances can be high for some time compared even with an institutionally good investment climate. In theory, organised corruption need not be detrimental to growth. It works like a "tax" for services metaphorised as "speed". The problem many a times is the unpredictable variability in the relationship between the two. No surprise that empirical assessments of the growth – corruption nexus offer a buffet of evidence: nonexistent, negative, positive, and nonlinear. Investor expectations are grounded in specific relationships to specific power bases. Shifts in power relationships precipitate sudden pivots towards better or worse as investor expectations realign to new realities. Divergence between the de jure laws and regulations and the de facto outcome for specific enterprises is evidenced by the resilience of informality alongside an expanding formal sector.
The typical degrees of regression to the mean imply substantial slowdowns following episodes of high growth in economies with very different rules of the game. Persistence of the relatively constant features of countries – culture, geography, the quality of institutions, openness to the world, climate – affect the level more than the long run growth of incomes.
Prosperity is no lottery
Given that Bangladesh experienced a period of accelerated growth in the past four decades, continued growth at 6-7% per year requires a relatively rare degree of persistence. Bangladesh appears to be completing a virtuous cycle of persistence that rose in the last half of the 90s and faded in the last half of 2010s after peaking in the second half of 2000s (Figure 1). The WB figures Bangladesh's annual growth rate for the coming decade could fall to 5% in less than two decades, absent renewed structural reforms.
As in Bangladesh, there is less persistence in the recent decadal growth rates globally than in previous decades. Summers and Pritchett report research identifying slowdown in countries with an episode of growth greater than 3.5% followed by a growth deceleration of 2% or more in the global income distributions of around Purchasing Power Parity adjusted $10,000-11,000 and another at PPP $15,000-16,000 per capita. These have been labelled as the "middle-income trap" turning economic miracles into a mirage.
Rapid growth associates with deceleration irrespective of a middle-income trap. Bangladesh is still distant from upper middle-income traps. The GDP per capita was PPP $6,020 in 2021, equivalent to 34% of the world's average and only 9.8% of the US level. However, even the most favourable conditions ultimately have diminishing impacts on growth unless a country continually tunes policy environment and governance to align with rise in savings, investments, and innovations.
Upgrading institutional quality is how the advanced economies in the East and the West got to where they currently are. Shifts in the rules of the game have long-lasting effects (Douglass North, 1990). The quality of institutions defines the constraints under which markets operate and preserve macroeconomic stability. Integration within and with the outside world has typically been a driver of productivity driven income growth. Less equal and cohesive societies typically experience more volatile growth.
Bangladesh navigated the pandemic relatively well. Bangladesh was initially hit harder by Covid-19 than most of its peers, except India. Economic activities recovered faster than in the peer countries and exceeded the pre-pandemic level by January 2021. Bangladesh entered the crisis with sufficient fiscal and monetary buffers while individuals and institutions rapidly adapted to the containment measures over time. However, resilience is more about building back better from shocks, not just endurance.
The capacity to manage risks will continue to be stressed. Growth in Bangladesh is currently facing the risk of global recession or even stagflation. At an 11-year high, inflation is dragging on economic activity and distributive justice amid deepening geopolitical uncertainty. The world appears edging toward a recession and a string of financial turbulence. The three largest economies – the US, China, and the Euro Area – have been slowing sharply. The risk of inflation remaining elevated for long, rising protectionism and fragmentation, while global growth is weak, looms large.
Meddlesome is cumbersome
Bangladesh's recent economic management has bit by bit drifted towards expanding government discretion vis-à-vis businesses and non-state actors. The embellishment of such an administrative state has added to the likelihood of a growth slowdown because of frictions that inevitably accompany bureaucratically driven economic management. Deepened conflicts of interest of regulators dealing with security of property rights and competition as well as with government-run financial and non-financial enterprises detract from the conditions that make for growth persistence.
Bangladesh's experience shows that a few years of market and competition enabling structural improvements can carry growth for a long period. Growing so fast for so long without new structural reforms is outside the general run of experience. The growing unexplained growth in recent years cautions against dismissing the need for structural adjustments based on the elusively "high" current growth. The regression to the mean may have been dwarfed by fuzzy statistics.
Meddling in the public availability of data compounds the existing gaps in knowledge. These gaps are reflected in the ongoing debates about the credibility of national accounts, price, reserves, financial asset quality, and trade statistics. Journeying safely to the Upper Middle-Income level is eminently achievable if policy and political will set the paddles to move forward at the needed speed by decisively shifting from centralised authority to credible rules-based economic policy frameworks.