ABSTRACT

International policy towards China faces inherent difficulty as a result of unreliability in official figures published by the Chinese state. This lessens the ability of foreign researchers and decision-makers to reach an accurate understanding of the Middle Kingdom. As a result, researchers have developed alternative methods to calculate figures like China’s real GDP growth rate. One of the most popular of these methods is the Li Index, a method used by Chinese Premier Li Keqiang in an admitted attempt to circumvent official GDP figures. Academic scholarship has found that the Li Index closely mirrors official Chinese GDP data. This paper confirms the findings of this previous scholarship and attempts to replicate the Li Index’s accuracy by applying it to other nations. This paper finds that the close correlation between the Li Index and GDP growth is only found in the Chinese data, and that when applied to other countries, the accuracy of the Li Index severely diminishes. This paper speculates that the cause of this divergence is likely either inaccuracy in the reported official Chinese data used to calculate the Li Index, or alternatively that the methodology used by the Chinese government to calculate their GDP figures places unequal weight on economic inputs in such a way as to privilege industrial figures.

1.     Background

As the world increasingly returns to the multipolar order that defined pre-20th Century history, the importance of China on the global chessboard will only grow with time. This growth has been and will continue to be accompanied by wariness on the part of China’s geopolitical rivals. Thus, a consistently growing segment of western policy research has been devoted to understanding the nature and actions of the Middle Kingdom. Unfortunately for scholars, this research is often clouded by the inherent lack of trustworthiness within official Chinese government data. The Communist Party of China (CCP) has spent decades building a reputation of statistical opacity and malfeasance which cloud the work of anyone who relies on the data they publish (Fernald, Malkin, & Spiegel, 2013).

This issue is deep-rooted in modern Chinese history, heavily intransient, and liable to cause trouble for policy-makers who rely on reports that use official figures. As such, many alternative methodologies have been proposed to measure traits such as Gross Domestic Product (GDP) in ways that more accurately reflect reality. For all their good intentions, however, these alternatives must not be accepted without scrutiny. Though no measurement conducted by human hands will ever be without flaw, certain systems of measurement and analysis are more likely to mirror reality, and this can only be determined through critical examination.

1.1.        Chinese Statistical Opacity

The use of national-level economic data aggregated by central states became a global fixture in the post-Bretton Woods world of the mid-20th Century. The World Bank, for instance, maintains a database of national GDP figures dating to 1960 (World Bank, 2022). This period coincides with the ascension of the CCP in 1949, serving as the practical birthdate of the modern Chinese state. Perhaps owing to its status as a Communist nation, the government of this new China mirrored its ally the Soviet Union both in the scientific trappings of its political philosophies as well as in the inaccuracy of its officially reported statistics.

One particularly illustrative example of this phenomenon is the “Illusion of Superabundance” (浮夸风) which served as a key building block of the 1959-1961 Great Chinese Famine. Often labeled as the most lethal famine event in recorded history, the Great Chinese Famine is estimated to have led to the starvation deaths of between 15 and 55 million people (Smil, 1999) and remains the single most lethal element of the infamous Great Leap Forward campaign. The Illusion of Superabundance emerged as a result of provincial governments over-reporting crop yields to the central government in Beijing. This over-reporting occurred due to the carrot-and-stick incentive structure of federal-local CCP relationships, wherein local political cadres would receive money and status for fulfilling arbitrary production quotas and would receive punishment for failure to meet the same quotas. In the absence of strict monitoring, it became a profitable risk for cadres to artificially inflate their province’s grain production numbers (Yang, 1998).

The consequences of this obfuscation were significant and deadly. Beijing’s agricultural redistribution programs would expropriate grain in such a way that enough was to be left behind to provide subsistence for provincial farmers. However, since the production numbers were hyperinflated, the amount taken away would often leave the provinces with little to no food at all, thus creating the conditions for starvation. Over-reporting posed another danger as well. As senior CCP officials were led to believe that agricultural production was at historic levels, the government felt safe in exporting large quantities of grain abroad for economic and political purposes, further stripping provinces of what little food they had available (Yang, 1998).

Inflation of agricultural figures in China continues to this day, though likely not to the same level, and clearly not to the same degree of lethality. In part, this over-reporting is due to the same carrot-and-stick incentive structure. However, modern agricultural obfuscation is often the result of individuals and local governments selling grain into the black market and simply reporting that silos remain as full as intended. Silos may be refilled with gravel or expired grain in order to maintain the illusion of the official numbers (Wanglei, 2021).

Though most infamous in the case of agricultural products, provincial over-reporting is also rife in the matter of local economic indicators, which are then used by the central government in Beijing to calculate national-level statistics like GDP. This was a noted issue in the days of Mao Zedong (Yang, 2012), and according to contemporary sources, this issue continues in the modern day as well (Fuxian, 2019). Researchers at the United States Federal Reserve have noted that China’s official national GDP numbers for 2018 were lower than the aggregate of all provincial GDP for that year by 7%. In essence, it appears as though the central government realizes that the provincial numbers they receive are overinflated and an unknown calculation is used to reduce the final GDP numbers to something that better reflects reality (Owyang & Shell, 2017).

This is not to say that China’s national GDP numbers are accurate by any means. Rather, western researchers generally hold the opposite view (Federal Reserve Bank of St. Louis, 2017). Untrustworthy local figures are only one reason for this lack of faith in the final product. Quite tellingly, the rate of annual Chinese GDP growth is much less volatile than those of other comparable nations. For instance, between 2006 and 2020, the Chinese GDP growth rate never changed by more than 458 basis points, while during the same period Singapore saw GDP growth rate swings as high as 1439 basis points. Even more suspicious, the CCP announces GDP growth rate targets early in the year, only for the final GDP figures to closely match what had been predicted nearly a year in advance (Owyang & Shell, 2017). This level of predictive power is evidence of either omniscience or manipulation.

1.2.      Proposed Methods for Uncovering Chinese GDP

The unreliability of Chinese GDP statistics presents something of a roadblock for foreign scholars and decision-makers. Though only one of many economic indicators – and not an infallible one at that – GDP serves as an easy-to-understand baseline from which comparisons of nations can be built. Neighboring Asian states look to China’s GDP numbers as a method of gauging regional dynamics. Adversarial states like the U.S. look to China’s GDP numbers to gauge relative strength. Developing states in regions like Africa and Central Asia look to China’s GDP numbers as an indicator of future foreign investment projects. Therefore, in the absence of reliable figures from the Chinese government, foreigners have taken upon themselves the task of creating proxy measurements which they hope can better approximate the reality of Chinese economic growth (Federal Reserve Bank of St. Louis, 2017).

In an oft-cited study by the University of Pittsburgh, Professor Thomas Rawski attempted one of the first published indexes of Chinese economic development by measuring the nation’s energy use, urban formal employment, and consumer price index. This Rawski Index claimed that official Chinese GDP growth rate figures were inflated by 5.6% annually in the late 1990’s (Rawski, 2001). As something of a spiritual successor to Rawski’s work, the Brown University Department of Economics published a 2012 study which sought to circumvent the inherent unreliability of international statistics by instead using satellite imaging technology to determine night-time luminescence as a proxy for a country’s total energy generation (Storeygard, Henderson, & Weil, 2012). Using this metric to correct for potential inaccuracies in Chinese data, it is estimated that total Chinese GDP growth between the years 1992 and 2006 was less than half of what was reported by official data (Owyang & Shell, 2017).

Several commercial research firms have developed proprietary indexes which they claim provide more accurate reflections of real Chinese economic growth than do the nation’s official statistics. These indexes include: the Barclays Chinese GDP Forecast, Bloomberg Intelligence’s Monthly GDP Tracker, Capital Economics’ China Activity Proxy, Lombard Street Research’s Real GDP Estimate, the Nomura Composite Leading Indicator, and the Oxford Economics GDP Proxy. These indexes utilize proprietary formulas which involve a blend of indicators that each respective firm believes acts as an economic proxy. Typical indicators include industrial output, retail sales, building units under construction, seaport cargo volume, and money supply. The accuracy of these indicators comes into question when looking at how significantly their results vary from one to the next. For instance, Barclays claims that official Chinese figures overstate GDP growth by roughly 1.5% each year, while Nomura claims that the official figures have been largely accurate. Lombard Street claims the official figures are overstated by 4.5%, while Bloomberg claims they are within the margin of error at only a 0.3% difference (Kawa, 2015).

Clearly, these various indexes – be they academic or commercial – cannot all be correct. Moreover, it cannot be definitively said which of these indexes are the most accurate, as all these indexes exist to solve a problem for which there is not an available answer. Perhaps, then, this confusion should be expected. After all, these various indexes come from foreign entities looking in. It is entirely possible that those metrics used by insiders of the Chinese system could offer more accuracy.

1.3.       The Li Index

The opacity of Chinese statistics limits the value of scholarship completed not only by western bodies, but also by the Chinese bodies who are obliged to use official government figures. Therefore, much like their foreign counterparts, individuals and organizations in China are apt to use alternative data and methodologies in order to reach a better approximation the truth than what is provided to them by their government.

The most famous of these domestic Chinese methodologies is a system referred to by researchers as the “Li Index.” This index derives its name from Li Keqiang, who recently served as the Premier of the State Council of the People’s Republic of China. The structure of the Chinese government is rather byzantine and is not easily mapped onto western systems of government, leading some commentators to refer to Li as China’s “Prime Minister” (Radio Free Europe, 2013). However, this title is not accurate. The State Council of the PRC more closely resembles an executive branch of government, as it encompasses all cabinet-level agencies and is responsible for the execution of state policy. The State Council is subservient to the National People’s Congress, a legislative body that determines state policy (People’s Daily, 2014). In real terms, this means that Li oversees the body responsible for the practical implementation of the grand strategies determined by the CCP. Thus, in official terms, Li is the second most powerful member of the Chinese government, behind only General Secretary Xi Jinping.

Li’s duty to execute lofty state policy in realistic ways is a task that suits his personality. By reputation, Li is known as a pragmatic man who eschews ideology in favor of practical results. Controversially within China, Li has been known to loudly contradict official statistics, such as in 2020 when he provided poverty rate numbers far afield of those reported by the state (PTI, 2020). Actions like this have earned Li the moniker of “reformer” from western press (Radio Free Europe, 2013), though his practicality should not be mistaken for political liberalism. Rather, Li is better compared to Deng Xiaoping, for former Chinese head of state who’s impressive – if limited – political and economic reforms helped bring China out of the shadow of Mao Zedong.

The 2020 incident was not the first time that Li had contradicted official state statistics. On March 15, 2007, while serving as the Communist Party Secretary of Liaoning Province, Li had dinner with the then-U.S. ambassador to China, Clark T. Randt. In a cable which Randt proceeded to send back to the U.S. State Department, he relayed comments that Li had made to him in confidence. In Li’s words – as relayed by Randt – “GDP figures are ‘man-made’ and therefore unreliable… When evaluating Liaoning’s economy, Li focuses on three figures: 1) electricity consumption, 2) volume of rail cargo, and 3) amount of loans disbursed… By looking at these three figures, Li said he can measure with relative accuracy the speed of economic growth.  All other figures, especially GDP statistics, are ‘for reference only,’ he said smiling.” (Randt, 2007).

In academic circles, this trifecta of measurements has become known – perhaps jokingly – as the “Li Index.” In the mind of the second most powerful man in China, the state’s official GDP figures are unreliable, and an index composed of electricity consumption, rail cargo, and loan disbursements provides a better look at the real rate of Chinese economic growth. Following the release of Randt’s cable by Wikileaks, the Li Index has become a niche topic of interest in western academic fields focused on the study of China, as it can be used as an alternative to China’s official GDP numbers. However, beyond the confidence of Li himself, it is difficult to know for certain if the Li Index is any more accurate than those indexes created by Bloomberg, Barclays, or any other institution. Casting even more confusion on this topic is a study conducted by the Federal Reserve Bank of San Francisco, who independently calculated their own version of the Li Index which they found to have high correlation to the official Chinese GDP statistics (Fernald, Malkin, & Spiegel, 2013).

2.     Analysis

This paper seeks to expand upon and compliment the work of the Federal Reserve in attempting to determine the accuracy of the Li Index. This paper proposes that the relative accuracy of the Li Index can be determined by applying the index not only to China, but also to a basket of various economies from around the world. In so doing, it is possible to tease out patterns that may either strengthen or weaken the case for Li Keqiang’s favorite statistical work-around.

2.1.      Calculating the Li Index

The Federal Reserve’s paper on the Li Index uses proprietary data on Chinese economic and industrial outputs. In order to maintain a level of consistency between the various states analyzed, this paper uses exclusively open-source figures provided by either national governments or major corporations, with an emphasis on finding the most comparable figures possible in an attempt to approximate an “apples-to-apples” comparison. GDP figures are drawn from the World Investment Report of the World Bank (World Bank, 2021). Electricity generation and consumption figures are drawn from the 2021 BP Statistical Review of World Energy, published by the former British Petroleum Company (BP Statistical Review Board, 2021). Loan disbursement figures are drawn from the Federal Reserve, which publishes separate reports by country (Federal Reserve Economic Data, 2022a) (Federal Reserve Economic Data, 2022b) (Federal Reserve Economic Data, 2022c) (Federal Reserve Economic Data, 2022d) (Federal Reserve Economic Data, 2022e) (Federal Reserve Economic Data, 2022f). Rail freight figures were the hardest to find in a consistent and reliable manner. For Chinese data, a series of Mandarin language reports are available open-source which provide annual figures between the years of 2006 and 2020 (Ministry of Transport of the People’s Republic of China, 2010) (Ministry of Transport of the People’s Republic of China, 2016) (Ministry of Transport of the People’s Republic of China, 2020) (Ministry of Transport of the People’s Republic of China, 2021). As such, this range of years has been selected for this paper’s analysis. Rail freight figures for the United States, Japan, and India are drawn from the open-source research organization Statista (Statista, 2022a) (Statista, 2022b) (Statista, 2022c). Rail freight figures for Germany are drawn from the European Statistical Office (Eurostat) (Eurostat, 2022).

2.2.     Li Index Correlation to Official Chinese GDP Figures

(Figure 1)

For the People’s Republic of China between the years of 2006 and 2020, the Li Index shares a correlation of 0.725 with the officially reported annual GDP growth figures. This is a statistically significant degree of correlation, which is in line with the results of the aforementioned Federal Reserve Study (Fernald, Malkin, & Spiegel, 2013). Taken in a vacuum, these results are significant in that they imply a level of reliability and accuracy to the official Chinese GDP statistics. Though these values and the Li Index do not perfectly reflect one another, the degree of correlation seems to strengthen the conclusion of the Nomura and Bloomberg indexes, which is to say that official Chinese figures are vindicated (Kawa, 2015).

However, the close correlation between the official figures and the Li Index begs the question of why the Li Index was necessary in the first place. If the difference between the official figures and the Li Index is relatively minimal, then why would Li Keqiang seek out an alternative method of measuring GDP growth? The two measurements are not identical, and so perhaps Li prefers the marginal difference of the Li Index as he believes it more closely reflects real economic conditions. However, the Li Index typically demonstrates a higher level of annual growth than do official GDP statistics. If, in fact, this relationship is correct, then it is strange that the Chinese government would comparatively understate the nation’s true GDP growth figures. Such consistent underreporting seems unlikely in light of known instances of significant over-reporting on the part of the Chinese state in the past.

2.3.     Li Index Correlation to Western States

Following the application of the Li Index to China and the confirmation of trends noted by the Federal Reserve, this paper applies the Li Index to other states in order to determine if a similar relationship exists. The United States and Germany were selected to serve as representative samples of major western nations. In particular, the U.S. was chosen for its role as China’s primary great power rival, and Germany was chosen as a representative European Union member state with an economy that is somewhat more reliant on industrial production than its peers, and thus more likely to be impacted by the indicators within the Li Index.

(Figure 2)

(Figure 3)

For the United States, the Li Index shares a correlation of 0.378 with the officially reported annual GDP growth figures. This is a statistically low-to-medium degree of correlation. For Germany, the Li Index shares a correlation of 0.240, which is a statistically low degree of correlation. These degrees of correlation are notable for how sharply they diverge with the Li Index’s Chinese correlation. If a high-ranking politician within the U.S. or Germany were to use the Li Index as an alternative to their government’s official figures, they would receive a degree of divergence much larger than that which Li Keqiang has historically calculated.

2.4.     Li Index Correlation to More Proximate States

It is possible that the degree of divergence in the Li Index’s accuracy between China and these two western states is a result of inherent differences between their respective economies. As a developing “middle income” state, China’s economic makeup may be naturally different from that of per-capita wealthier nations, and perhaps it is this difference which explains the degree of divergence.

As such, the remainder of the nations analyzed by this paper were chosen in an attempt to approximate elements of the Chinese economy that might have led to such a high Li Index correlation. Japan is a geographically and genetically proximate nation which experienced a similar degree of explosive economic development in the mid-20th Century. Both nations are also highly reliant on rail transport. India, in addition to being a major geopolitical rival of China, is the only nation on Earth with a comparably large population and shares a similar rural-urban economic divide. Both nations also share a similar relationship of exports as a percent of GDP (18.5% for China, 18.7% for India) (World Bank, 2022). Finally, Singapore was chosen for its similar consumer market size (39% of GDP for China, 36% for Singapore) (World Bank, 2022). Moreover, Singapore’s largest and most economically powerful ethnic group are the Han, descendants of Chinese emigrees, leading to high ethic and cultural proximity between the two nations. However, due to a lack of open-source data on Singaporean rail freight passage, the Li Index as used for Singapore has been modified to only incorporate electricity consumption and loan disbursement.

(Figure 4)

For Japan, the Li Index shares a correlation of -0.416, which is a statistically low-to-medium degree of negative (inverse) correlation. This is a startling degree of difference from the correlation found in the Chinese data. Clearly, the sample of Japan fails to validate the Li Index as an accurate measure of economic growth. In fact, the Li Index is more apt to predict the opposite of real GDP growth rate when applied to Japan.

(Figure 5)

For India, the Li Index shares a correlation of 0.252, which is a statistically low degree of correlation. This degree of correlation is comparable to that seen in the example of Germany, with both serving as datapoints which fail to validate the Li Index’s accuracy.

(Figure 6)

For Singapore, the Li Index shares a correlation of 0.383, which is a statistically low degree of correlation. This degree of correlation is comparable to that seen in the example of the United States, which additionally fails to validate the accuracy of the Li Index, although to a lesser degree than the examples of Japan, Germany, and India.

2.5.     Limits of Analysis

There are various factors and experimental variables which may limit the accuracy of these results, or which may be responsible for the discrepancy discovered between China and the other nations observed. For instance, this paper applies equal weight to each of the three elements which compose the Li Index. It is possible that Li Keqiang himself uses a different weighing ratio when he applies his analysis, but it is rather unlikely that Li will share this ratio with interested foreign parties. Additionally, it is possible that the open-source data used in this analysis is inaccurate as a result of intentional or unintentional errors on the part of various reporting agencies. If, for instance, the nations of the U.S., Germany, Japan, India, and Singapore were to have inaccurately reported their economic data – and if the government of China alone had reported their data correctly – this could serve as an explanation for why the statistical gap is so significant between China and the other states.

However, given the preponderance of historical evidence, this paper speculates that the wide gap between the Chinese Li Index results and the international Li Index results is more-likely-than-not the result of data issues in the Chinese datasets. Applying Occam’s Razor, if Chinese data is incongruent with data from other major nations, it is most likely that the Chinese data is inaccurate.

3.     Conclusions

3.1.       Data Malfeasance

Stated simply, the Li Index is an accurate indicator for Chinese GDP figures if one trusts both the Chinese GDP figures and the Chinese figures which compose the Li Index. The accuracy of both these sets of figures is thrown into doubt when the Li Index is applied to other nations, where none of the states tested demonstrated a Li Index correlation anywhere close to that found in China.

The results of this paper alone are not sufficient to claim with certainty why these discrepancies exist. However, these results are sufficient enough to allow for speculation of possible explanations. The most obvious of these possible explanations is that economic data reported by the Chinese government is inaccurate. This applies to GDP figures, of course, but also to the economic indicators used to create the Li Index. It does not make logical sense for Li Keqiang to use the Li Index if it provides results commensurate to the same official GDP figures which he distrusts. Rather, it seems likely that Li has access to figures for electrical consumption, rail freight transport, and loan disbursement which are different from those provided to the general public.

Serving recently as the Premier of the State Council – and serving previously as the Party Secretary of Liaoning Province – Li is likely one of the few men in China with the power and authority to procure accurate, high-level economic data. In the leaked Randt cable, Li states that he uses his chosen three figures because they can be definitively verified. For instance, Li stated that he focuses on rail freight volume “because fees are charged for each unit of weight,” and he focuses on loan disbursements because it “also tends to be accurate given the interest fees charged” (Randt, 2007). This can be interpreted to mean that Li is able to use his position to independently verify electricity, rail freight, and loan numbers, rather than having to rely on those provided by the government. Of course, for foreign researchers without access to Li’s power, it is much more difficult to verify the accuracy of the economic data that has been made public by the Chinese state.

3.2.     Unequal Statistical Weight

Let us speculate for a moment further and assume that China’s officially reported electricity, rail, and loan data are all accurate. A discrepancy still remains in the Li Index-GDP growth relationship found in China and the same relationship found in other nations. Therefore, if the end result of a calculation does not match, and if the inputs of the equation are all accurate, then the only remaining explanation is that the calculation itself is incorrect. If Chinese economic data is accurate, and if the final GDP relationship is incongruent with the rest of the world, then the formulation of China’s GDP data must be different than those used by other nations. It is possible that the government in Beijing calculates their final GDP figures in such a way as to place greater statistical weight on industrial factors like electricity consumption, rail freight volume, and loan disbursement.

There is circumstantial evidence that may support this theory. The Li Index as calculated for China is often higher than the annual GDP growth rate. If Chinese physical industry grows at consistently higher rates than other elements of the economy, then the Chinese state could artificially boost their GDP figures by simply weighing GDP calculations more towards industry. We know conclusively that the CCP makes revisions to their GDP numbers thanks to the discrepancy discovered between regional and national GDP reports (Owyang & Shell, 2017). It is not outside the realm of possibility that they would make similar revisions to inflate figures in a flattering way.

3.3.     Lessons for Foreign Analysts

Foreign researchers and decision makers are advised to remain cautious of official Chinese statistics. When analyzed comparatively with economic data from other nations, Chinese economic data produces strange and incongruitous results. Unfortunately, in the absence of more leaked information from sources within the CCP, the exact nature of how and to what extent this data is unreliable still cannot be known. With that in mind, a healthy skepticism is recommended.

It is also recommended that this skepticism be extended to indexes and formulations that claim to be able to decipher Chinese data. The various commercial and academic attempts to “fix” China’s reported data have led to significantly different results, and thus far none of these indexes have demonstrated their accuracy above the others. This sentiment applies equally to the Li Index, which should theoretically be the most accurate of indexes as it is used by the head of China’s executive branch. It is possible that the Li Index is of great value when inputted with accurate data, but foreign analysts have no way of knowing if their input factors are accurate when taken from official sources.

The Li Index faces further challenges as a result of China’s changing economic landscape. The use of three, very industry-weighted data points as a replacement for GDP is a tactic that could only theoretically work in a simple, undiversified economy. Li’s three favorite factors are reminiscent of the Maoist economic era, where the nation’s prosperity was boiled down to a small handful of production targets (Yang, 1998). As China grows increasingly modern, increasingly digital, and increasingly diversified, whatever accuracy the Li Index once had will become diminished with time.

While a one-stop-shop index is an enticing proposition, it is ultimately of limited use to researchers and decision-makers. When approaching difficult, entrenched issues like international policy towards China, a more wholistic approach is required. The only one universal truth to be applied in such circumstances is to maintain caution and to maintain skepticism. The black box of the Chinese state is particularly opaque, and true wisdom is often found in admitting just how little we know.

REFERENCES AND BIBLIOGRAPHY

BP Statistical Review Board. (2021, July). bp Statistical Review of World Energy July 2021. BP p.l.c.. http://www.bp.com/statisticalreview

EuroStat. (2022, September 29). Goods transported by type of transport. https://ec.europa.eu/eurostat/databrowser/view/RAIL_GO_TYPEPAS__custom_3471754/default/table?lang=en

Federal Reserve Bank of St. Louis. (2017, September 12). Some Alternative Methods for Tracking Chinese GDP. https://www.stlouisfed.org/on-the-economy/2017/september/some-alternative-methods-tracking-chinese-gdp

Federal Reserve Economic Data. (2022a, September 19). Amount Outstanding of International Debt Securities for Issuers in Non-Financial Corporations (Corporate Issuers), All Maturities, Residence of Issuer in India. FRED. https://fred.stlouisfed.org/series/IDSNFAMRIAOIN

Federal Reserve Economic Data. (2022b, September 19). Amount Outstanding of Total Debt Securities in Non-Financial Corporations Sector, All Maturities, Residence of Issuer in China. FRED. https://fred.stlouisfed.org/series/TDSAMRIAONCCN

Federal Reserve Economic Data. (2022c, September 19). Amount Outstanding of Total Debt Securities in Non-Financial Corporations Sector, All Maturities, Residence of Issuer in Germany. FRED. https://fred.stlouisfed.org/series/TDSAMRIAONCDE

Federal Reserve Economic Data. (2022d, September 19). Amount Outstanding of Total Debt Securities in Non-Financial Corporations Sector, All Maturities, Residence of Issuer in Japan. FRED. https://fred.stlouisfed.org/series/TDSAMRIAONCJP

Federal Reserve Economic Data. (2022e, September 19). Amount Outstanding of Total Debt Securities in Non-Financial Corporations Sector, All Maturities, Residence of Issuer in Singapore. FRED. https://fred.stlouisfed.org/series/TDSAMRIAONCSG

Federal Reserve Economic Data. (2022f, September 19). Amount Outstanding of Total Debt Securities in Non-Financial Corporations Sector, All Maturities, Residence of Issuer in United States. FRED. https://fred.stlouisfed.org/series/TDSAMRIAONCUS

Fernald, J., Malkin, I., & Spiegel, M. (2013, March 25). On the Reliability of Chinese Output Figures. Federal Reserve Bank of San Francisco. https://www.frbsf.org/economic-research/publications/economic-letter/2013/march/reliability-chinese-output-figures/

Fuxian, Yi. (2019, July 20). China’s Population Numbers are Almost Certainly Inflated to Hide the Harmful Legacy of its Family Planning Policy. South China Morning Post. https://www.scmp.com/comment/opinion/article/3018829/chinas-population-numbers-are-almost-certainly-inflated-hide

Kawa, Luke. (2015, November 2). Six Ways to Gauge How Fast China’s Economy Is Actually Growing. Bloomberg. https://www.bloomberg.com/news/articles/2015-11-02/six-ways-to-gauge-how-fast-china-s-economy-is-actually-growing?leadSource=uverify%20wall

Ministry of Transport of the People’s Republic of China. (2010). 2010 年铁道统计公报. ChinaGate. http://cn.chinagate.cn/reports/2012-09/11/content_26494227_3.htm

Ministry of Transport of the People’s Republic of China. (2016). 2015 年铁道统计公报. https://www.mot.gov.cn/tongjishuju/tielu/201510/t20151016_1906182.html

Ministry of Transport of the People’s Republic of China. (2020). 2019 年铁道统计公报. https://www.mot.gov.cn/tongjishuju/tielu/202005/P020200702602040585281.pdf

Ministry of Transport of the People’s Republic of China. (2021). 2020 年铁道统计公报. https://www.mot.gov.cn/tongjishuju/tielu/202205/P020220507531780768964.pdf

Owyang, Michael. & Shell, Hannah. (2017, July 25). China’s Economic Data: An Accurate Reflection, or Just Smoke and Mirrors? Federal Reserve Bank of St. Louis. https://www.stlouisfed.org/publications/regional-economist/second-quarter-2017/chinas-economic-data-an-accurate-reflection-or-just-smoke-and-mirrors

People’s Daily. (2014, November 14). The State Council. http://en.people.cn/data/organs/statecouncil.shtml

PTI. (2020, May 28). China Has Over 600 Million Poor with $140 Monthly Income. CNBC. https://www.cnbctv18.com/economy/china-has-over-600-million-poor-with-140-monthly-income-premier-li-keqiang-6024341.htm

Radio Free Europe. (2013, November 7). Profile: Chinese First Vice Prime Minister Li Keqiang. https://www.rferl.org/a/china-profile-li-keqiang/24764285.html

Randt, Clark. (2007, March 15). Fifth Generation Star Li Keqiang Discusses Domestic Challenges, Trade Relations with Ambassador. Wikileaks. https://wikileaks.org/plusd/cables/07BEIJING1760_a.html

Rawski, Thomas. (2001, January). What is Happening to China’s GDP Statistics? China Economic Review, 12(4), 347–354. https://doi.org/10.1016/s1043-951x(01)00062-1

Smil, Vaclav. (1999, December 18). China’s Great Famine: 40 Years Later. British Medical Journal, 319(7225), 1619–1621. https://doi.org/10.1136/bmj.319.7225.1619

Statista. (2022a, May 12). Volume of Freight Carried by Indian Railways FY 2010-2021. https://www.statista.com/statistics/741325/india-rail-freight-volume/

Statista. (2022b, August 1). Domestic Railway Freight Volume Transported in Japan FY 2011-2020. https://www.statista.com/statistics/628262/japan-size-railway-freight-transport/#:%7E:text=In%20the%20fiscal%20year%202020%2C%20railway%20freight%20in,rail%20cargo%20volume%2C%20freight%20cars%20for%20the%20rest.

Statista. (2022c, September 28). Rail Freight Volume Transported in the U.S. 2002-2020. https://www.statista.com/statistics/245235/rail-freight-transportation-in-the-us/

Storeygard, Adam., Henderson, J. Vernon., & Weil, David. (2012, April 1). Measuring Economic Growth from Outer Space. American Economic Review, 102(2), 994–1028. https://doi.org/10.1257/aer.102.2.994

Wanglei, Zhou. (2021, June 26). Henan Farmer Who Denounced Grain Storage Scandal Arrested and Tortured. Bitter Winter. https://bitterwinter.org/henan-farmer-who-denounced-grain-storage-scandal-arrested-and-tortured/

World Bank. (2021). World Development Indicators. https://databank.worldbank.org/source/world-development-indicators#

Yang, Dali. (1998, August 1). Calamity and Reform in China: State, Rural Society, and Institutional Change Since the Great Leap Famine (1st ed.). Stanford University Press.

Yang, Jisheng. (2012, October 30). Tombstone: The Great Chinese Famine, 1958-1962. Farrar, Straus and Giroux.