1 Introduction
Hong Kong was Milton Friedman’s “favorite economy,” as he believed it was the “freest” in the world with minimal government involvement. This idea of a free market with minimal government interference was first suggested by John Cowperthwaite, the colony’s Financial Secretary, in the 1960s. The city’s Financial Secretaries in the decades since have followed this golden rule without challenge. Sixty years after Friedman’s visit to Hong Kong, the city is a key financial center in the Far East and has the second most billionaires in the world after New York, according to Forbes in 2023. It is also one of the world’s richest cities, with a GDP per capita of $49,800 in US dollars as of 2021.
However, alongside this economic success, Hong Kong faces significant social challenges. Its housing market is the most expensive in the world. The high cost of housing has resulted in a homeownership rate of only 51.5% in 2022, compared to 89.3% in its competing sister city, Singapore. An estimated 220,000 people live in subdivided flats with living spaces as small as 30 square feet. Among them, approximately 34,000 are children[1]. Behind Hong Kong’s housing inequality lies high income inequality, with the Gini coefficient of household income surpassing that of both San Francisco and London. This disparity has been suggested as one of the primary factors contributing to the major social unrest witnessed in Hong Kong in 2019.
Despite regular media coverage of income inequality in Hong Kong, there is a lack of comprehensive analysis on the city’s cross-sectional inequality, including wages, individual earnings, household income, and government transfers. In this paper, we provide a systematic and multifaceted analysis of economic inequality based on Hong Kong Census data from 1981 to 2016, covering approximately 94,000 households per wave, or 5% of total households. Our data allows us to explore inequality trends before and after Hong Kong’s 1997 handover. Such discussion is important because excessive income inequality can erode social cohesion and lead to political polarization and social instability, making it a significant public policy concern.[2]
See related discussions on inequality by IMF (https://www.imf.org/en/Topics/Inequality/introduction-to-inequality) and the Brookings Institution (https://www.brookings.edu/articles/rising-inequality-a-major-issue-of-our-time/).
In this study, we aim to identify the changes in patterns of individual wages and household earnings in Hong Kong over the past 25 years. We find that the growing wage inequality may be driven by high-income earners, who contribute to higher overall household inequality. Despite the rising median household income, our findings suggest that the income of the poorest households is regressing. Together, these findings paint a picture of economic inequality in Hong Kong. However, there are some surprising findings as well. We observe greater gender wage equalization and find that government transfers have played a more significant role since the handover of Hong Kong. We also find that the wage gap between high-skilled and low-skilled workers have declined in Hong Kong, due to the government’s investment in higher education that resulted in large expansions of college enrollment. These findings suggest that without government intervention, inequality in Hong Kong may be even higher. The idea of minimal government interference may no longer hold true in the area of income distribution in this city.
2 Historical and Institutional Background
Before the Korean War in the 1950s, Hong Kong served as an entrepot. According to Sung’s (1986) estimation based on historical records of the Hong Kong government, approximately 80% of the imports from mainland China to Hong Kong were re-exported prior to the Korean War. This volume represented 16% of China’s total exports at that time. However, the role of re-exportation began to decline after the United Nations imposed an embargo on trade with China in 1951.
In response to the decline of Hong Kong’s entrepot role, the city’s manufacturing industries experienced rapid growth during the 1950s and 1960s. The influx of manufacturers and capital from mainland China during the civil war provided the necessary know-how and resources for industry development. In his classic work on emigrant entrepreneurs from Shanghai in Hong Kong, Wong (1988) depicted how the Shanghai textile industry reestablished itself in Hong Kong after 1949, becoming the cornerstone of the Hong Kong economy in the 1950s and 1960s. The industry expanded from 132,000 spindles in 1950 to 900,000 in 1970. Additionally, related industries such as toy and watch production experienced rapid growth.
The growing manufacturing base and the increased demand for a large volume of low-paid unskilled workers led many refugees fleeing China to fill these positions. This classic scenario of industrialization contributed to income inequality, with a clear division between labor and capital. Partly due to poor working and housing conditions among the workers, as well as the rampant corruption within the city during its rapid industrialization, and partly influenced by the Chinese Cultural Revolution (Yep 2023), major civil unrest erupted in Hong Kong in 1967. This unrest resulted in significant violence, leading to 51 deaths and close to 5,000 arrests.
Sir Murray MacLehose’s November 1971 arrival as the new governor marked the start of government efforts to rebuild the colonial city following the 1960s unrest. MacLehose introduced policies that laid the foundation for the thriving Hong Kong economy in the decades to come, including implementing nine years of compulsory education and a ten-year housing program, establishing of the Independent Commission Against Corruption, and creating satellite new towns such as Sha Tin and Tuen Mun. In the 1980s, Hong Kong, along with Singapore, South Korea, and Taiwan, became collectively known as the Four Asian Tigers.
With the onset of economic reforms in China in the late 1970s, jobs in Hong Kong gradually shifted to mainland China as foreign investment was encouraged. Simultaneously, the expansion of education in Hong Kong during the 1980s led to an increase in the proportion of the labor force that had completed higher education. After Hong Kong’s return to mainland China in 1997, there was an influx of highly educated migrants from the mainland seeking economic opportunities. The supply of highly educated labor exceeded the demand, leading to limited returns on higher education.
Income inequality in Hong Kong was accompanied by gender income disparity in the 1960s and 1970s. As the center of manufacturing industries, a large number of low-paid female workers were recruited, as described by Salaff (1995) in her well-known account of female factory workers in Hong Kong during the 1970s. This employment pattern exacerbated gender inequality. However, the growth of service industries in the 1980s created a large number of low-paying jobs for both males and females. Fan and Lui (2003) demonstrated that the transformation from a manufacturing to a service-oriented economy, coupled with a lower gender wage gap in services, helped narrow the gender income gap in Hong Kong.
The narrowing gender income gap was further reinforced by legislation as society became more aware of gender-related issues. The Equal Opportunities Commission, established in 1996, immediately implemented the Sex Discrimination Ordinance of 1995. The Commission played a vital role in raising awareness of gender issues and handling complaints of gender discrimination. Female workers have benefited from this increased social awareness, resulting in a considerable reduction in gender income inequality.
Although the Hong Kong government has adhered to a laissez-faire policy, it has made efforts to assist those in poverty through various programs. The “Comprehensive Social Security Assistance” (CSSA) program was initiated in 1993 to provide support to those in need. This program was renamed from the previously established Public Assistance program, which began in 1971. Over the years, the amount of support and eligibility criteria of the CSSA program have been revised to adapt to changes in the economy. In addition to this program, the government has implemented the Working Family Allowance Scheme, which provides allowances to households based on the parents’ working hours, as well as a monthly allowance for each child. Furthermore, schemes such as the Kindergarten and Child Care Center Fee Remission Scheme and the Grant for School-related Expenses for Kindergarten Students have been introduced to assist families in need. Various assistance schemes also provide support for children in primary, secondary, and tertiary education, including funding for continuing education. These programs have been implemented to support households with limited resources. However, critics argue that the support provided to households has been limited.
3 Measuring Inequality in Hong Kong
Our paper uses Hong Kong Census/By-Census data (Census data), which collects 1 percent (before 1991) or 5 percent (after 1991) samples of the total population in Hong Kong. We include eight waves of Census data: 1981, 1986, 1991, 1996, 2001, 2006, 2011, and 2016. The basic unit of observation is a housing unit, so we report Census statistics on inequality at the household level rather than the family level. On average, there are nearly 94,000 households per wave (except 1981, 1986). Census data includes detailed demographic characteristics for each household member and income variables for individuals aged 15 or older.
Following the methodology from Heathcote, Perri, and Violante (2010) and Heathcote, Perri, Violante, and Zhang (2023), we use the household budget constraint as an organizing device and construct different samples for analyzing income inequality at different levels.[3]
An appendix on data and method is available from the authors upon request.
We start with individual wages. Then we incorporated individual labor supply to study earnings, added other household members to analyze household earnings, and moved to broader income measures by including unearned income (e.g. rent, interest, and dividend income), private transfers, and government transfers. Our data enables us to explore inequality trends before and after Hong Kong’s 1997 handover. We summarize our five key findings as follows.
3.1 Wage inequality widens and highest earners pull away
Monthly income from primary employment offers a simple yet significant perspective on inequality, as wages and salaries constitute the largest income component for most Hong Kong residents, excluding a small fraction of the ultra-wealthy.
Figure 1 illustrates an overall increase in wage inequality from the early 1990s to 2011, as evidenced by the variance of log and Gini coefficient, irrespective of gender. A distinct pattern emerges when comparing the top, middle, and bottom earners. The wages of the top 10% have significantly risen, particularly during the 1980s for men and the 1990s for women. However, this growth has slowed and even reversed in recent years, with the highest earners now making approximately 5.5 times the median. Conversely, the wage gap between median and low-income workers has remained relatively stable, though more volatile for women, over the past four decades.
3.2 Greater gender equalization
In line with global trends documented by 2023 Nobel Laureate Claudia Goldin in Goldin (2006), we also observe greater gender equality among Hong Kong workers. While many men’s earnings have stagnated over the past 40 years, women’s labor market outcomes have improved. Women have considerably narrowed the gap with men in both monthly wages and labor supply in terms of total employment share (Figure 2).
In 1981, the average working woman earned an hourly wage 60% that of the average working man, and women’s total employment share was around 30%. By 2016, these figures had increased to approximately 74% and 50% respectively, although the pace of progress has slowed since the mid-2000s.
3.3 Wage premium for college graduates has declined
In most advanced economies, the skill premium – the wage premium for employed workers with college degrees relative to those without college degrees – has significantly grown since the 1980s. However, the skill premium in Hong Kong decreased significantly since the early 1990s. The decline is due to a cohort effect: Within a cohort, the skill premium increases over time, but newer cohorts start with a lower skill premium when they enter the labor force. (See Figure 3). Two potential explanations for this finding exist from the supply and demand sides of the high-skill labor market.
On the supply side, two higher education expansions in 1989 and 2001 resulted in dramatic growth of gross tertiary education enrollment. According to available statistics (UNESCO, 2021), the Hong Kong gross tertiary education enrollment rate grew from 10.12% in 1980 to 80.98% in 2019. Figure 4 shows that the share of college-educated workers increased across cohorts, especially after cohort 1965. The increasing supply of high skill workers helps explain the downward pressure on wages for skilled workers.
On the demand side, while the structural change from manufacturing to services has helped reducing gender wage gap, it has not generated enough demand for high skill workers. Figure 5 shows the value-added shares of high-skilled, low-skilled, and real estate services for Hong Kong, Japan, mainland China, and Singapore. Hong Kong’s share of high-skilled services is comparable to that of Singapore, but low-skilled and real estate services account for much larger shares than their counterparts in other economies. Figure 6 shows the corresponding employment shares that exhibit similar patterns. Thus, high reliance on low-skilled and real estate services in Hong Kong has limited its demand for high skill workers.
3.4 Median household income is rising, while poorest households regress
When individuals with different wages live together, sharing income and expenses, their combined earnings contribute to the household’s standard of living. Economists use a standard methodology to “equivalize” the earnings of different-sized households for comparison purposes (Figure 7). This approach reveals that median household earnings have more than doubled since 1981, after accounting for inflation.
Households in the bottom 10% have experienced less progress, with earnings declining, particularly during the 1997 Asian Financial Crisis (light blue curve in Figure 7). For the poorest households, real earnings have decreased over the past 40 years.
In contrast, households near the top have seen earnings rise between 2 to 3 times. However, this measure does not account for non-labor income from financial holdings or business ownership.
Strikingly, despite societal advancements in various dimensions over the past four decades, households in the bottom 10% of Hong Kong’s income distribution now earn even less than they did 40 years ago.
3.5 Government transfers plays a more important role after the Handover of Hong Kong
Atop wages and household income, the government provides transfers, such as food stamps and discounted services for low-income Hong Kong people. We include these transfers in our household income measure, creating a pre-tax income measure.[4]More precisely, pre-tax income = Earnings + other cash income (asset income + private transfers + government transfers). Due to data limitations, we cannot separate government transfers from the other two components of the category on other cash income. However, we believe that most of the other cash income should be government transfers since private transfers are generally a small component, and asset income is usually poorly measured in survey-based data. Figure 8 displays the evolution of household income inequality in terms of the 90th percentile to the 20th percentile ratio in household earnings distribution with and without considering government transfers. Comparing the household earnings inequality (blue solid curve) with the household pre-tax income inequality (red dashed curve) reveals the significant role government transfers play in reducing inequality. Further analysis of household income inequality evolution at the top and bottom (Figure 9) confirms that government transfers have a more considerable impact on reducing inequality at the lower end.
This becomes even more evident in Figure 10 when we calculate transfers as a share of pre-tax income for the top 10%, middle 10%, and bottom 20% in the household income distribution. Before the 1997 handover, the transfer shares in the three groups of households were surprisingly similar. However, after the handover, there was a surge in transfers received by the bottom group, with the transfer share increasing from around 15% to almost 40%.
The key takeaway is that after the 1997 handover, although earnings inequality increased significantly, more government transfers or benefits were provided to lower-income households, which considerably reduced income inequality at the bottom. This conclusion differs from Piketty and Yang (2021), who measured income inequality based on wages and salaries without considering government transfers.
4 Conclusion
In this study, we have used comprehensive micro data to measure income inequality in Hong Kong and examine its trend over the last 25 years. We find a growing income inequality that is driven by faster income growth of high-income earners. Despite the rising median household income, our findings suggest that the income of the poorest households is regressing. We also find that government transfers and investment in higher education have played significant roles in reducing poverty and wage inequality between high-skilled and low-skilled workers, respectively. These findings suggest that without government intervention, inequality in Hong Kong may be even higher. We also note that over reliance on low-skilled and real estate services in creating jobs has limited job opportunities and income growth of younger generations of college graduates in Hong Kong. Finally, we should emphasize that due to high price of housing, wealth inequality are likely to be significantly higher than the income inequality we have measured in this study.
Therefore, to reduce inequality in Hong Kong, the government should invest more in public housing and promoting job creation in high-skilled services by reducing entry barriers.
References:
Fan, Chengze Simon, and Hon-Kwang Lui. “Structural Change and the Narrowing Gender Gap in Wages: Theory and Evidence from Hong Kong.” Labor Economics, 10 (5), (2003), pp.609-626.
Heathcote, Jonathan, Fabrizio Perri, and Giovanni L. Violante. “Unequal We Stand: An Empirical Analysis of Economic Inequality in the United States, 1967–2006.” Review of economic dynamics 13, no. 1 (2010): pp.15–51.
Heathcote, Jonathan, Fabrizio Perri, Giovanni L. Violante, and Lichen Zhang. “More Unequal We Stand? Inequality Dynamics in the United States, 1967–2021.” Review of economic dynamics 50 (2023): pp.235–266.
Piketty, Thomas, and Li Yang. “Income and Wealth Inequality in Hong Kong, 1981–2020: The Rise of Pluto-Communism?.” The World Bank Economic Review 36, no. 4 (2022): pp.803-834
Salaff, Janet. Working Daughters of Hong Kong: Filial Piety or Power in the Family. Columbia University Press, (1995).
Sung, Yun-wing. “The Role of Hong Kong in China’s Export Drive.” The Australian Journal of Chinese Affairs, no. 15, (1986), pp. 83–101.
Wong, Siu-Lun. (1988). Emigrant Entrepreneurs: Shanghai Industrialists in Hong Kong. Oxford University Press, Hong Kong.
Yep, Ray. “The 1967 Riots in Hong Kong: The Diplomatic and Domestic Fronts of the Colonial Governor.” The China Quarterly, no. 193, (2008), pp. 122–39.
Figure 1: Wage Inequality
Note: X-axis is the survey year from 1981 to 2016. Gini coefficient measures the extent to which the distribution of income within a country deviates from a perfectly equal distribution. P50-P10 ratio compares the monthly wage at the 50th percentile to the one at the 10th percentile. P90-P50 ratio compares the monthly wage at the 90th percentile to the one at the 50th percentile.
Figure 2: The greater gender equalization
Note: X-axis is the survey year from 1981 to 2016. Gender wage gap refers to the ratio of female average monthly wage to male average monthly wage. Employment share refers to the share of employed females/males in total employed population.
Figure 3: Skill Premium
Note: X-axis is the survey year from 1981 to 2016. “Cross-sectional” includes all samples, and “Cohort-Level” separates the sample into 5 birth cohort groups
Figure 4: Share of Highly-Educated Workers by Cohort
Note: X-axis is the birth cohort from 1926 to 1988. Y-axis is the ratio of the number of workers with college degree to the total employed population by each cohort.
Figure 7: Evolution of household earnings distribution
Note: X-axis is the survey year from 1981 to 2016. Household earnings refer to the sum of all individuals’ annual earnings in each household. Each line represents the trend of average household earnings by earnings percentiles and normalized to 0 in 1981.
Figure 8: Household earnings inequality v.s. pre-tax income inequality
Note: X-axis is the survey year from 1981 to 2016. P90-P20 ratio compares household earnings/pre-tax income at the 90th percentile to the one at the 20th percentile.
Figure 9: Household earnings inequality v.s. pre-tax income inequality: top and bottom
Note: X-axis is the survey year from 1981 to 2016. P50-P20 ratio compares household earnings/pre-tax income at the 50th percentile to the one at the 20th percentile. P90-P50 ratio is defined as in Figure 1.
Figure 10: The distribution of government transfers
Note: X-axis is the survey year from 1981 to 2016. Y-axis is the average percentage of other income in household pre-tax income.
[1] Source: https://www.scmp.com/yp/discover/news/hong-kong/article/3235398/hong-kong-set-open-community-living-room-residents-subdivided-flats-years-end.
[2] See related discussions on inequality by IMF (https://www.imf.org/en/Topics/Inequality/introduction-to-inequality) and the Brookings Institution (https://www.brookings.edu/articles/rising-inequality-a-major-issue-of-our-time/).
[3] An appendix on data and method is available from the authors upon request.
[4] More precisely, pre-tax income = Earnings + other cash income (asset income + private transfers + government transfers). Due to data limitations, we cannot separate government transfers from the other two components of the category on other cash income. However, we believe that most of the other cash income should be government transfers since private transfers are generally a small component, and asset income is usually poorly measured in survey-based data.
