Why per capita GDP cannot be the benchmark for minimum wage
In a recent Facebook post, Nazmul Ahasan, executive editor of Netra News, argued that comparisons between the minimum wages of garment workers in Bangladesh, Vietnam, and China are misleading when used to advocate for higher wages in Bangladesh.
His core argument is that Bangladesh's minimum wage is already relatively high when measured against the average income of the population. He points out that Bangladesh's minimum wage of 12,500 taka represents approximately 42% of per capita income, compared with 27% in China and 33–47% in Vietnam.
From this, he concludes that a demand for a 30,000 taka minimum wage is economically unrealistic and that garment workers in Bangladesh are relatively well paid.
His argument rests on three problems: conflating GDP with income, using national averages as benchmarks for minimum wages, and failing to consider the cost of living.
GDP is not income, and income statistics exist
Let us accept the premise that minimum wages should be linked to average income and examine the methodology. Although Ahasan repeatedly refers to “per capita income”, the figures he cites are per capita GDP. These are not interchangeable concepts.
GDP measures the total value of goods and services produced within a country's geographical boundaries. Per capita GDP is this figure divided by the country's population. Average income, by contrast, refers to the average income earned by individuals or households, depending on how it is measured. The distinction matters because not all economic output generated within a country necessarily accrues to its residents.
For example, profits generated by foreign-owned businesses operating in Bangladesh contribute to the country's GDP. However, when those profits are repatriated abroad, they do not become income retained by Bangladeshi residents. This illustrates why GDP and residents' income are different concepts. In Bangladesh, average income is around half its per capita GDP ($1,600 versus $2,900).
After some people pointed out that median income is a more appropriate measure than average income, Ahasan suggested that median income data was unavailable, which was why he resorted to average salaries. His original comparison, however, relied on GDP per capita, not average salaries.
The claim that median income data is unavailable is inaccurate. The International Labour Organization's ILOSTAT database contains labour income and wage statistics drawn from nationally representative Labour Force Surveys, among other sources. The Labour Force Survey 2024 of Bangladesh is also publicly available.
While the ILO website's main interface displays average income figures, it provides a link to a more detailed dataset containing both average and median values across demographic groups. Ahasan's claim that median income data is unavailable, particularly after the relevant dataset had been pointed out to him, suggests that he may not have explored the full dataset before making the claim.
The availability and comparability of specific indicators may vary across countries and survey years. But that differs from claiming that the data does not exist. For anyone making an argument about labour economics, distinguishing between GDP and income and understanding the availability of relevant data are fundamental. Failure to do so raises questions about credibility.
Average income is not a meaningful benchmark for minimum wages
Consider a hypothetical country in which 90% of workers earn exactly the minimum wage, while the remaining 10% earn significantly more. In such a country, the minimum wage would equal the median wage and be relatively close to the average wage.
Now imagine that the government doubles the minimum wage, with the increase financed, for simplicity, by a reduction in income for the top 10%. Workers' nominal earnings would double, potentially improving their living standards. Yet the relationship between the minimum wage and the median wage would remain unchanged: both would rise together. Would that mean nothing had changed for workers? No.
This illustrates the limitation of using a ratio between the minimum wage and a measure of central tendency (average or median) as an indicator of wage adequacy. Such a ratio tells us where the minimum wage sits in relation to the income distribution. It does not tell us whether workers can afford a decent standard of living.
Income distributions are heavily right-skewed. In most economies, a relatively small share of high earners pulls the average upwards, while the majority earn less. This is why the mean and median can tell different stories. A proportional comparison between a minimum wage and a national average can therefore be misleading, particularly when comparing countries with different income distributions.
Per capita GDP is further removed from the question. It includes economic output that does not necessarily translate into workers' disposable income, and it conceals how income is distributed across the population. Consequently, showing that Bangladesh's minimum wage represents a larger proportion of per capita GDP or average income than China's does not establish that Bangladeshi garment workers are better paid relative to their economic circumstances. Nor does it establish that a higher minimum wage would be unreasonable.
An established economic indicator for comparing minimum wages with average or median wages is the Kaitz Index. It is widely used by the ILO, OECD, IMF, and other institutions to assess the relative position of minimum wages within national wage distributions.
For example, Colombia's Kaitz Index is around 90%, while the OECD average is approximately 50%. Does this mean that workers in Colombia enjoy a better standard of living than workers in OECD countries? No. It indicates that Colombia's minimum wage is much closer to the median wage than is the case in the OECD on average.
In the United States, for example, the national minimum wage is low relative to the median wage, producing a Kaitz Index of around 25%. If the federal minimum wage were raised to $15 an hour, as many US states have already done and as many people support, the index would move considerably closer to the OECD average.
The distinction matters. Comparing minimum wages with average or median wages tells us where minimum-wage earners stand relative to other workers. It does not tell us whether their earnings are sufficient to meet their basic needs. That requires a separate assessment of purchasing power and the cost of living.
Ahasan's analysis conflates these two questions. Even if his calculations were corrected to compare the Kaitz Index across these countries, they could not establish that Bangladesh's garment workers are adequately paid or that a higher minimum wage is economically unjustified.
Minimum wages should reflect the cost of living, and Bangladesh falls significantly short
The purpose of a minimum wage is not to maintain a particular statistical relationship with national income. It is to ensure that workers receive enough to meet their basic needs. Workers must be able to afford food, housing, healthcare, transport, clothing, and other essentials for themselves and their dependents.
If the minimum wage cannot cover these costs, workers are left to endure an inadequate standard of living, work excessive overtime, or depend on government or private charity for assistance.
This is why the relationship between minimum wages and the cost of living is a more relevant measure of wage adequacy. International comparisons must also account for differences in living costs. According to the World Bank's Price Level Index, the overall price level of household consumption in Vietnam was identical to that of Bangladesh in 2025 (an index of 28.9), while China's was higher (48.1).
Yet garment workers' minimum wages in both countries are substantially higher than in Bangladesh: almost twice as high in Vietnam and around three times as high in China.
Aggregate price-level indices are no substitute for detailed household expenditure surveys. Living costs also vary between cities and regions within the same country. Nevertheless, these indicators demonstrate why comparisons based solely on minimum-wage-to-average-income ratios are insufficient to assess workers' economic circumstances.
The relevant question is: how much does a worker need to earn to maintain a basic standard of living, and how far does the existing minimum wage fall short of that amount? In Bangladesh, the gap is substantial.
Concluding Remarks
There is a legitimate debate about the appropriate minimum wage for Bangladesh's garment workers. The industry's financial capacity, productivity, employment effects, international competitiveness, and distribution of profits all deserve consideration. But that debate must begin with appropriate economic indicators and an understanding of basic economic concepts.
Per capita GDP is not the same as income. The ratio of the minimum wage to average or median income does not establish whether workers are adequately compensated. And comparisons that ignore the cost of living cannot tell us whether workers in different countries enjoy comparable living standards.
Ahasan argues that 30,000 taka is unrealistic because it would amount to approximately 101% of Bangladesh's per capita GDP. But this calculation, by itself, tells us little about whether 30,000 taka is an appropriate minimum wage.
The more relevant question is whether workers can afford a decent standard of living on their wages.
Economic growth and international competitiveness should not be assessed independently of the conditions under which workers produce that growth. Economic progress must not depend on the exploitation of workers. Bangladesh's garment industry has become a major global exporter. Ensuring that the people who sustain this industry can afford a decent life should be part of evaluating its economic success.
--
The writer is a PhD candidate and the information and research secretary of the National Citizen Party (NCP).