Theoretical literature on drivers of household savings in China tends to focus on the interaction between household consumption trends and overall savings. China’s high household savings rate has been attributed to demographic factors, the one-child policy, precautionary savings amidst a weak social safety net, weak financial services availability, and rising income inequality. In a cross-country empirical analysis of developed and developing countries, Loayza, Schmidt-Hebbel, and Serven (2000) find private and national savings are affected by the level of per capita income, economic growth, fiscal policy, pension reform, financial liberalization, and demographics.78 The life-cycle hypothesis developed by Modigliani and Brumberg in the 1950s emphasizes factors such as a country’s economic growth, growth prospects, and demographic structure, in particular of the proportion of the workforce in the population, rather than the level of income in making saving decisions.79 The main implication of the hypothesis is that the national saving rate is unrelated to per capita income but instead depends on the long-term rate of income growth.
Rodrik (2000) similarly argues high saving rates tend to be the outcome of high rates of growth but not a determinant of faster growth.80 The permanent income hypothesis, formulated by Milton Friedman, posits that consumption is insensitive to transitory fluctuations in income that do not affect permanent income in the longer run.81 This argument would imply consumer spending, and therefore savings levels, need not move together with current income levels.
76 Reuven Glick and Kevin Lansing, “Consumers and the Economy, Part 1: Household Credit and Personal Saving” Federal Reserve Bank of San Francisco Economic Letter 01, January 10, 2011, https://www.frbsf.org/economic-research/publications/economic-letter/2011/january/consumers-economy-household-credit-personal-saving/.
77 Xie Yu and Jin Yongai, “Household Wealth in China,” China Sociological Review, vol. 47, no. 3, (2015), 203-299, https://doi.org/10.1080/21620555.2015.1032158.
78 Norman Loayza et al., “What Drives Private Saving around the World?” Policy Research Working Paper, no. 2309, World Bank, March 2000, p. 18, https://openknowledge.worldbank.org/bitstream/handle/10986/18854/multi_page.pdf.
79 Franco Modigliani and Richard Brumberg, “Utility analysis and the consumption function: an interpretation of the cross-section data,” in Kenneth Kurihara, ed. Post-Keynsian Economics (New Brunswick, NJ: Rutgers University Press, 1954), 388-436.
80 Dani Rodrik, “Saving Transitions,” World Bank Economic Review, vol. 14, no. 3, (1998), 483, http://faculty.ucr.edu/~jorgea/econ181/rodrik_wber2000.pdf.
81 Friedman, Milton, A Theory of the Consumption Function, National Bureau of Economic Research (1957).
Demographic factors are significant in understanding China’s high household saving rate and tend to focus on the impact of the one-child policy. According to the IMF, the rapid decline in the fertility rate and the relative decline in the youth population increased Chinese household savings in two ways.82 First, the working age population engaged in higher precautionary savings to prepare for retirement, as there are fewer children to become the primary caretakers of retirees in China. Secondly, fewer children required less total spending. In a 2013 study, Choukhmane, Coeurdacier, and Jin quantify the impact, finding that the one-child policy accounts for 30 to 60 percent of the rise in aggregate household savings since its implementation.83 Extending the life-cycle hypothesis to the case of China, Modigliani and Cao (2004) estimate two factors contributed equally and significantly to the rise in China’s saving rate in the post-reform period: income growth ignited by the market-based post-reforms of the 1970s; and demographic
changes which both reduced the number of youth within the working population and undermined the role of children in providing old-age support, prompting more savings.84 Inflation largely accounted for the remainder of growth in the household saving rate from the 1970s to 1994.85 Likewise, Curtis et al. (2012) find the change in population composition alone generates over half of the observed increase in household saving from 1955 to 2009.86 This echoed the findings of Wang Wei of the Shanghai University of Finance and Economics in 2009, who argued that starting in the 1970s, family planning profoundly impacted savings rates because they significantly reduced birth rates and cut family expenditures because there were fewer mouths to feed.87 Another demographic factor contributing to higher savings includes a rise in the labor force participation rate to 63 percent in 2006 from 42 percent in 1979, which structurally increased the amount of income that could be saved.
China’s household savings rate has also risen in part because of inadequacies in the social safety net, which leads to precautionary savings. Within the last decade, household savings has become far more concentrated at the top of the income spectrum, with income inequality a key factor contributing to China’s household savings, as wealthier households tend to consume a smaller proportion of their income.88 Fan Jianjun, a researcher with the Development and Research Center of the State Council, emphasized inadequacies in China’s social security scheme that overcovered some people while providing inadequate coverage for others, therefore keeping savings rates high and making it difficult to narrow the income gap.89 Hu Cui of the Central University of Finance and Economics and Xu Zhaoyuan of the Development Research Center of the State Council also argue the different social security schemes in urban and rural China led to differential impacts of population aging on the savings rate.90 They argue that urban Chinese were better covered by insurance so they did not need to save as much to prepare for future spending (although this finding is contrary to the growing concentration of savings at the top of the
82 “People’s Republic of China: Selected Issues,” International Monetary Fund, IMF Country Report, no. 17/248, August 2017, p. 7.
83 Taha Choukhmane, Nicolas Coeurdacier, and Keyu Jin, “The One-Child Policy and Household Savings,” Working Paper, August 24, 2013.
84 Franco Modigliani and Shi Larry Cao, “The Chinese Saving Puzzle and the Life-Cycle Hypothesis,” Journal of Economic Literature, vol.
XLII, (March 2004), 166-167, http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.186.42&rep=rep1&type=pdf.
85 Ibid.
86 Chadwick Curtis et al., “Demographic Patterns and Household Saving in China,” Working Paper, no. 006, University of Notre Dame, June 2012, 20, ftp://ftp.repec.org/opt/ReDIF/RePEc/nod/wpaper/006_saving.pdf.
87 Wang Wei, “Economic Growth, Demographic Transition and China’s High Savings,” China Economic Quarterly, vol. 9, no. 1, (October 2009).
88 See, for example, Chu Tianshu and Wen Qiang, “Can Income Inequality Explain China’s Saving Puzzle?” International Review of Economics and Finance, vol. 52, (November 2017), 222-235; Thomas Piketty, Li Yang, and Gabriel Zucman, “Capital Accumulation, Private Property and Rising Inequality in China, 1978-2015,” HKUST IEMS Working Paper, no. 2018-54, (March 2018),
http://iems.ust.hk/assets/publications/working-papers-2018/iemswp2018-54.pdf.
89 Fan Jianjun, “Analyzing the Reason for China’s Relatively High Savings Ratio,” Development Research, no. 10, (October 2014).
90 Sun, Hanlin, “Empirical Analysis on the Relation Between Population Aging and Resident Consumption in China: Based on the Data From 1981 to 2011,” Management Science and Engineering, vol. 8, no. 2, (2014), 79-88.
income spectrum). Therefore, the savings rate did not fall along with an aging population as traditional theory would have suggested. However, in rural areas, where social security systems have not been fully developed, aging does tend to bring down the savings rate. The authors concluded that the ongoing pace of urbanization meant that China’s overall savings rate would not decline because of an aging population.91 Literature over the past decade has placed more emphasis on the impact of income inequality on China’s household savings rate, alongside the relatively weak levels of government spending on transfer payments.
Savings rates measured by different income deciles within China highlight these trends. According to data from the Chinese Household Income Project (CHIP), a joint survey of households conducted by
international and Chinese economists, the poorest decile of China’s population saved just under 20 percent of their annual income—significantly higher than a survey of nine other countries, where savings rates are negative in the poorest decile.92 Urban household savings rates have shifted higher over the past three income surveys, with a profound shift from 2007 to 2013. The savings ratio according to income decile shows that the lowest 10 percent of households saved 20 percent of their take-home pay while the richest 10 percent saved close to 50 percent, which is less surprising given the rapid growth in wealth in China relative to the slower rise in overall consumption. In comparing China’s situation with other countries—in France, Taiwan, and Australia—all but the top four deciles have negative savings rates.
China’s substantial positive savings rates across the entire income spectrum signal underdeveloped fiscal institutions such as taxation (specifically an adequate and progressive income tax), social safety nets, and other transfer payments.93 A survey of household finances conducted by the Southwestern University of Finance and Economics reinforces the role of income inequality in promoting high savings in China. This survey finds that the average savings rate of the top 10 percent of households is about 66.6 percent, comprising an astounding 75 percent of total national savings.94 The same survey also found that the top 5 percent of households by income had average savings rates as high as 69 percent and represented a total of 62 percent of total household savings.95 This echoed earlier work from Zhang Ming of the Chinese
Academy of Social Sciences, who noted in a 2005 paper that the top income quintile held over 60 percent of China’s household savings.96 Jin Ye, Li Hongbin, and Wu Binzhen, a team of researchers and professors from Tsinghua University, put forth a more novel argument on the impact of wealth inequality. Using the urban household survey data, they found income inequality has a negative impact on consumption and, therefore, predictably boosted the savings ratio.97 However, they also argued individuals saved to improve their social status, and more inequality meant more required savings to attain social mobility, which strengthened the incentives to save.
However, the literature on China’s savings rates still leaves several questions unanswered, particularly the impact of many years of income growth on China’s household savings rates. According to traditional savings models, prolonged double-digit income growth in China would result in individuals saving less and consuming more under reliable expectations of income growth. In 2000, for example, Carroll, Overland,
91 Hu Cui and Xu Zhaoyuan, “Empirical Analysis of the Population Ageing and Saving Rate: The Data From Chinese Families,” China Economic Quarterly (经济学(季刊), vol. 13, no. 4 (July 2014), 1345-1364.
92 “People’s Republic of China: Selected Issues,” International Monetary Fund, IMF Country Report, no. 17/248, August 2017, p. 9.
93 Ibid.
94 Gan Li, “Findings from China Household Finance Survey,” Texas A&M University and Southwestern University of Finance and Economics, January 2013, http://people.tamu.edu/~ganli/Report-English-Dec-2013.pdf
95 Ibid.
96 Zhang Ming, “A Perspective on the High-Saving Phenomenon in China: Efficiency Loss and Factor Analysis,” Shanghai Economic Review, no. 8, (August 2005).
97 Jin Ye, Li Hongbin, and Wu Binzhen, “Income Inequality, Status Seeking and Consumption,” China Economic Quarterly, vol. 10, no.
3, (April 2011).
and Weil introduced the idea of a “habit stock” of consumption, in which savers use past, rather than future, consumption as a frame of reference to determine savings rates.98 Contrary to this model, household savings rates in China have risen the most among households with relatively younger and relatively older heads, whereas typically savings rates would increase until peaking prior to the household head’s retirement. Chamon, Liu, and Prasad (2010) argue the main cause is a form of “self-insurance”
against rising income uncertainty. They use a buffer-stock savings model which estimates that, after the breaking of the “iron rice bowl” of publicly-provided education, health, and housing services, rising income uncertainty and declining pension replacement rates contributed to about half the increase of the urban household savings rate since the middle of the 1990s.99 However, Choi, Lugauer, and Mark cast doubt on this theory by constructing a model using principles similar to Carroll, Overland, and Weil’s in a 2017 study. The authors posit that households have a target wealth-to-income ratio, implying that higher income growth drives higher precautionary savings to maintain this ratio.100 In comparing precautionary savings between U.S. and Chinese households, the study finds that the difference in savings rates are almost entirely determined by varying growth rates, and that if GDP growth in China slowed to 2 to 3 percent, aggregate savings would also drop by around 10 percentage points.101
Figure 3-6: Household Income Growth and Savings, 1992-2017 Year-on-year change (LHS); Percent of GDP (RHS)
Source: National Bureau of Statistics.
Overall, the fact that many years of rapid income growth have not reduced China’s household savings rates remains inconsistent with most of the conventional literature on savings rates. Demographic factors and precautionary savings certainly play a role but have probably not significantly changed over the past
98 Christopher Carroll, Jody Overland, and David Weil, “Saving and Growth with Habit Formation,” American Economic Review, vol. 90, no. 3, (June 2000), 341-355.
99 Marcos Chamon, Kai Liu, and Esawar Prasad, “Income Uncertainty and Household Savings in China,” International Monetary Fund, December 2010, https://www.imf.org/en/Publications/WP/Issues/2016/12/31/Income-Uncertainty-and-Household-Savings-in-China-24497.
100 Horace Choi, Steven Lugauer, and Nelson Mark, “Precautionary Savings of Chinese and U.S. Households,” Journal of Money, Credit, and Banking, vol. 49, no. 4, (June 2017).
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Urban Household Disposable Income Growth (Survey) Household Total Disposable Income Growth (Flow of Funds) Household Savings (% of GDP, RHS)
decade, while household savings rates have remained high. However, the recent literature discussing the impact of China’s income inequality probably best explains the recent rise in savings rates, as well as the concentration of household savings at the very bottom and the very top of the income spectrum, with most household savings concentrated at the top (and possibly invested in the property market as well).
This suggests that any attempt to change household savings behavior in aggregate would probably involve changing the incentives of the wealthiest Chinese households to save in some form.