The villain is winning!

The detailed Household Income and Expenditure Survey (HIES) 2022 data, made public recently by the Bangladesh Bureau of Statistics (BBS), carries good and bad news. The good news is both the incidence and depth of poverty was significantly lower in 2022 relative to 2016. The bad news is the pace of poverty reduction is anaemic relative to the pace of observed income growth. Increasing inequality is the villain who seems to be beating the hero called economic growth.

Poverty decline is broad-based

Most of the received literature focuses on socio-economic conditions of the population as mediators of the relationship between growth and monetary measures of poverty reduction. BBS provides data on important correlates such as geographic location, household size, literacy, occupation, and land ownership.

Poverty reduction between 2016 to 2022 happened across all regions, unlike in the previous rounds of surveys in 2016 and 2010. The reduction in rural poverty was 3.7 times higher than in urban areas. The highest incidence of poverty (26.9%) is in the Barisal Division and the lowest in Khulna Division (14.8%). Rangpur Division has the second highest poverty rate (24.8%) followed by Mymensingh Division (24.2%). The poverty rate in Dhaka (17.9%), Sylhet (17.4%), Rajshahi (16.7%), and Chattogram (15.8%) divisions are in between.

Poverty reduction during 2016-2022 happened across all household sizes. The direct relationship between poverty and household size is robust across surveys. The Head Count Ratio tends to increase with household size. However, this relationship is not exactly linear. National incidence of poverty at the upper poverty line for HIES 2022 is lowest (6.8%) for households with 1-2 members and highest (29.2%) for households with 7-8 members.

Poverty reduction cuts across households at all educational levels. Historically, illiteracy has been associated with a higher incidence of poverty. The national poverty rate at the upper PL is 26.9% for illiterate individuals compared to 14.2% for literate individuals. Poverty rates for illiterate individuals are close in rural (27%) and urban areas (26.2%), while poverty among literate individuals in rural areas (16%) is higher than in urban areas (11.1%). Poverty rates decline with the level of education in all HIES. However, the pace of HCR decline was remarkably faster for illiterate compared with literate during 2016-2022.

Poverty declined in most occupations. Service workers were the poorest in 2022 with a poverty rate of 22.6% using the upper PL. At the lower PL, the poorest individuals are employed in the "Production, Transport and Related sector" with a poverty rate of 8.4%. In rural areas, "Service Workers" have the highest incidence of poverty at 25.5% at the upper PL. "Agriculture, Forestry & Fisheries" occupations in urban areas have the highest 22.7% poverty rate.

An inverse relationship between land ownership and poverty incidence is a historic regularity. Land-poor individuals have higher rates of poverty. This relationship is alive and well in HIES 2022. The incidence of poverty generally decreases with increase in the size of land holdings. However, the decline in poverty happened across all land groups during 2016-22 at a noteworthy pace.

There was a broad-based reduction in poverty across geography, household size, education, occupation, and land ownership which points to the role of factors beyond structural correlates in explaining poverty reduction. Income growth and its distribution are central in this regard.

Low sensitivity to aggregate income growth

The pace of poverty reduction is weak. Backward calculations of poverty based on the reconstructed 2022 poverty lines results in an upward adjustment of the 2016 poverty estimate based on the Upper Poverty Line from 24.3% to 26.4% and a downward adjustment in the extreme poverty rate estimate from 12.9% to 9.3%. These imply 1.3 percentage point annual reduction in poverty and 0.6 percentage point annual reduction in extreme poverty during 2016-2022. Hard to say how these compare with the previous survey periods because the 2010 poverty estimate is strictly not comparable with the revised 2016 estimate.

Income growth underpins poverty reduction because the measurement of poverty incidence is anchored on expenditure. Most studies point to considerable heterogeneity in the poverty-growth nexus. Economic growth has a better chance at reducing poverty when income inequality is stable over time. Exactly how much better depends on the rate of economic growth itself and the extent of inequality.

Bangladesh's growth elasticity of poverty, the ratio of the percentage decline in poverty to any measure of the percentage decline in income, is low compared to peer countries. It was 1.02 during 2016-2022 with respect to per capita real GDP growth and 1.54 with respect to per capita GNI (in Atlas US dollars) growth. The difference between the two is attributable to lower GNI per capita growth. Standard estimates of Growth Elasticity of Poverty for developing countries range from 1.5 to 5, with an average of around 3.

The responsiveness of poverty to economic growth is heavily contingent on a whole variety of other factors. Salient among those is inequality in the distribution of income. Countries with a more equal distribution of income experience a greater reduction in the poverty rate for a given increase in per capita income. In other words, for any given rate of economic growth, the more inequality falls, the greater is the reduction in poverty. Even small changes in the overall distribution of inequality can lead to sizeable changes in the incidence of poverty.

You do not have to look outside Bangladesh for evidence in this regard. The Gini coefficients of income and consumption were respectively 3.5% and 3.1% higher in 2022 relative to 2016. All of the increase in income and consumption inequalities occurred in urban locations where the poverty reduction was also lower (0.7 percentage point per year). In rural areas where income inequality decreased by 1.8% during 2016-2022, the rate of poverty reduction was faster (1 percentage point per year).

Note that nominal income, expenditure, and consumption growth were all lower in rural compared to urban locations irrespective of whether measured on household or per capita basis. Yet rural poverty declined faster in terms of percentage points per year and by an equal 4.1% annually despite lower income, expenditure, and consumption growth!

Inequality morphing into a major development barrier

No matter what measure of income inequality you look at, it is becoming a serious problem in Bangladesh. Bangladesh's income Gini coefficient rose to 0.5 in 2022 at the national level. There is no definitive verdict on what value of the income Gini coefficient is too high or too low. The Gini coefficient only measures the dispersion of income or wealth within a population. A high-income country and a low-income country can have the same Gini coefficient. However, based on experience, it's commonly recognised that Gini index less than 0.4 corresponds with a relatively reasonable income gap and one above 0.5 corresponds with severe income disparity.

The Palma Ratio, another measure of income inequality, focuses on the extremes of the income distribution and captures the relative gap between the rich and the poor. Considered more robust than the Gini coefficient, this ratio typically compares the income share of the richest 10% to the income share of the poorest 40%. In Bangladesh, the PR increased from 2.9 in 2016 to 3.2 in 2022. This means the richest 10% now have 3.2 times the income share of the poorest 40%. The richest 10% accounted for 41% of the country's total income in 2022, compared with 38% in 2016.

Inequality in Bangladesh is heading in the wrong direction. The Palma ratio is not widely available for all countries. Available data suggest it is highest in South Africa, Namibia, Mexico, Honduras, and Brazil (6 to 7) and lowest in Finland, Norway, Ukraine, Slovenia, and Belarus (1 to 1.5).

Poverty and inequality are non-separable

There is an increasing recognition of poverty and inequality as two sides of the same coin even though the relationship between the two is not yet well understood theoretically and empirically. Income inequality and income poverty have followed similar trends in many countries. Many researchers find a positive correlation between the levels and change in income inequality and various measures of poverty. These do not look like mere statistical coincidences. There may be several economic, social, and political mechanisms driving the observed correlations.

At what level does income inequality pose a socioeconomic risk is essentially an empirical question. Some studies suggest that a Gini coefficient above 0.4 may have negative effects on economic growth, social cohesion, political stability, human development, and environmental sustainability. Additionally, inequality reduces the growth potential of an economy, limits the opportunities and capabilities of the poor, and undermines social cohesion and trust.

Income disparities are partly the result of unequal pay for work and partly of much larger inequalities in income from capital. These, at the end of the day, are themselves consequences of the extreme concentration of wealth. Growing income inequality is little explained by the tranquil intersectoral mobility often used by the policy makers (popularly known as the Kuznets Curve) to make inequality sound like an indicator of economic progress. Evidence is growing in favour of the argument that strategies aimed at reducing poverty without reducing inequality are nonstarters.