The measurement of inequality and poverty is crucial for informing social welfare policies. However, methodologies for these measurements have been debated since the 1970s. The Gini Index, a widely adopted measure of inequality, focuses on relative deprivation, reflecting income differences from the population median. Similarly, relative poverty measures define poverty as income below a certain percentage of the median income. While relative measures are favored for developed economies, capturing changing needs over time and contextual factors, they are criticized for neglecting absolute needs and failing to capture social mobility. The author presents evidence from studies in China, showing that absolute and relative poverty lines miss significant portions of the impoverished population. This article argues for a cost-of-living approach, emphasizing the need to consider the actual cost of essential goods and services when determining poverty lines, instead of relying solely on relative income comparisons. The high reliance on service-based workforce in Hong Kong makes it an ideal case for comparing the two approaches.

The article reviews the literature on inequality and poverty measurement, highlighting the debate between absolute and relative measures. It discusses the limitations of the Gini index, which is insensitive to changes in income distribution that are symmetrical at both ends of the spectrum, and thus fails to accurately reflect shifts in social mobility. The author critically examines the conventional relative poverty measures that define the poverty line as a percentage of median income, arguing that this approach is arbitrary and susceptible to underestimating poverty during economic downturns, particularly when median income decreases alongside rising cost of living. The author cites various studies that highlight the inadequacies of both absolute and relative measures in accurately capturing the extent of poverty, especially in developing and transitioning economies. The article also points out the influence of political ideologies in the preference for relative measures, as evidenced by the popularity of slogans such as the "99% vs. 1%" narrative.

The study uses Hong Kong as a case study to compare relative and cost-of-living approaches to measuring inequality and poverty. The Gini index is used as the measure of inequality, along with two formulas for calculating the poverty rate. The first poverty rate calculation uses the conventional relative measure, defined as the percentage of households with income below 50% of the median household income (Equation 3). The second calculation adopts a cost-of-living approach. The author constructs a cost-of-living index (Equation 5) by tracking the prices of nine essential categories of goods and services in Hong Kong from 1996 to 2020. These categories include food, housing, clothing, durable goods, miscellaneous goods, transportation, and miscellaneous services. The index is then used to determine the poverty line in Equation 6, representing the real cost of essential goods and services. Data sources include the Census and Statistics Department of Hong Kong, the Legislative Council of Hong Kong, and the World Bank. The author analyzes changes in poverty rates and poverty lines under both approaches, particularly during periods of economic crisis in Hong Kong.

The study reveals that the Gini index, even when decomposed into subgroups by household size, masks significant shifts in household income and the number of households in different income brackets. Relative measures of poverty, as shown in Figure 4, are found to underestimate the actual extent of poverty in Hong Kong. During economic crises, the poverty line based on the median income shows a lag in adapting to rising prices and unemployment rates, making it unreliable during these times (Table 2). In contrast, the cost-of-living approach (Figure 5) paints a more accurate picture of poverty, indicating a much higher poverty rate than the relative measure. The cost-of-living approach calculates a poverty line of HK$28,815 in 2020, significantly higher than the relative measure's HK$13,450. The resulting poverty rate in 2020 from the cost-of-living approach (44.47%) is almost double the 23.6% reported from relative measure, showcasing a difference of 551,400 poor households (Figure 6).

The findings highlight the limitations of relying solely on relative deprivation measures for assessing poverty and inequality. The cost-of-living approach offers a more accurate and nuanced understanding of poverty by focusing on the actual cost of essential goods and services. The significant discrepancy between the poverty rates derived from relative and cost-of-living approaches demonstrates the need for a shift towards more robust measurement methods. The Gini index’s limitation of failing to differentiate between inequality at lower and higher income percentiles and the relative measure’s lag in capturing changes in economic means during crises emphasize the need to consider both absolute needs and relative income when measuring poverty, aligning with Ravallion's observation. The cost-of-living approach is presented as a more suitable alternative, which considers material needs and is less prone to underestimating poverty during economic uncertainties. This approach shows that poverty is not just a function of income distribution but also the material needs that define a minimum standard of living.

This study demonstrates that relative measures of poverty, such as those based on median income, significantly underestimate the true extent of poverty. The cost-of-living approach offers a more accurate alternative by grounding poverty lines in the real costs of essential goods and services. The study highlights the importance of considering the inherent limitations of commonly used metrics and the need for more robust and context-specific approaches. Further research could explore the dynamics of income elasticity of different needs, adjustments to household purchases after their income falls below the poverty line, and the interconnection between inequality and poverty in diverse economic contexts.

The study’s use of a household size equivalent of 2.5 persons as a baseline for the cost-of-living calculation, similar to the Hong Kong government’s approach, may introduce limitations as the distribution of household sizes changes. The cost-of-living approach does not entirely capture the elasticity of substitution of products within a spending module, nor does it consider the adjustments made by households to their consumption patterns when faced with income constraints. Furthermore, the study focuses on a particular urbanized context in Hong Kong and may not generalize to economies with a larger share of agricultural or rural workers. Finally, the study does not directly investigate the relationship between inequality and poverty.