Before dawn on 19 May 1989, Zhao Ziyang, General Secretary of the Chinese Communist Party, walks into Tiananmen Square.
It is the seventh day of the student hunger strike, and by the New York Times’s count about 3,000 students are fasting. The protests began in mourning. Hu Yaobang, the former General Secretary, died on 15 April, and on 26 April the People’s Daily called the protests a planned conspiracy. Martial law will be declared on 20 May.
He tells them that he and his colleagues have come too late. He urges them to end the fast, and he says that dialogue will stay open. Things are complicated, he says, and there must be a process to resolve them.
It is his last public appearance. The next day Premier Li Peng signs the martial-law order. Zhao had opposed it, and Deng Xiaoping took Li’s side. Zhao will spend about fifteen years under house arrest. In 1999 and 2000 he will record some thirty cassettes of memories, and they will be published in 2009 as Prisoner of the State.
Two weeks later, on the night of 3 June and the early hours of 4 June, the troops advance. The figures that survive in the open record come from US embassy cables. They recorded estimates that began at 180 to 500 dead, then rose to between 500 and 2,600 dead with up to 10,000 injured. On 22 June the embassy passed on a Chinese Red Cross figure of 2,600 dead and 7,000 wounded, and called it not an unreasonable estimate. It judged that most of the deaths came on Chang’an Avenue and the streets near it, not in the Square. A claim of more than 10,000 dead was never verified. The book keeps these as a range from named documents and does not pick a number. On 9 June, Deng addressed the commanders who had enforced martial law. He called the events a counter-revolutionary rebellion, and he said that reform and opening must continue. The Fourth Plenum, on 23 and 24 June, removed Zhao from all his posts and made Jiang Zemin General Secretary.
The growth rate shows the cost of the winter and the speed of the recovery. China’s economy grew 4.2 percent in 1989 and 3.9 percent in 1990, down from 11.2 percent in 1988, and then 9.4 percent in 1991. India grew 6.0 percent in 1989, 5.5 percent in 1990 and 1.1 percent in 1991. Foreign investment into China barely moved in the clampdown, at $3.4 billion in 1989 and $3.5 billion in 1990.
India, that autumn, is changing its government at the ballot box. In two years it will have had three Prime Ministers: Rajiv Gandhi, V. P. Singh and Chandra Shekhar. In November 1989 Congress won more seats than any other party but not a majority, Rajiv Gandhi resigned on 29 November, and a coalition under V. P. Singh took over. Chapter 10 tells how it fell, and what the two years cost. This chapter asks what each country’s way of choosing its rulers has to do with the other country’s boom.
A reader who has watched both years may feel a familiar argument forming. In The Brothers Karamazov, Ivan tells his brother a story set in Seville. Christ returns, and the aged Grand Inquisitor has him arrested and visits him in his cell at night. Freedom, the old man says, is a burden most people cannot bear. Give them bread, and they will lay their freedom at your feet. In Constance Garnett’s translation he asks the question as a battle cry.
Feed men, and then ask of them virtue!
The Grand Inquisitor · Dostoevsky, The Brothers Karamazov · tr. Constance Garnett (1912)
It is hard to look at modern China without hearing that voice. After 1989 the Party’s unwritten offer is often described in nearly those terms: rising prosperity, in exchange for leaving politics to the Party. Deng himself said on 4 March that tighter control would not frighten off foreign investors and might reassure them. The bill for the bread was real: the one-child limit reached into every family, and chapter 3 tells how, and the household registration that decides where a migrant’s child may go to school is told in chapter 17. The tempting conclusion, heard at dinner tables from Bengaluru to Boston, is that India chose freedom and paid in bread; that democracy is simply slower; that to get China’s growth you need China’s state.
That conclusion needs one thing to be true: that the rulers’ power and the people’s bread move together. Three scenes from 1947 to 1949, and the famine that followed, say they do not.
Three scenes from 1947 to 1949 show how each new state first met its people.
On 15 August 1947, the day India became independent, Mohandas Gandhi was in Calcutta. He had arrived two days before and was staying at Hydari Manzil, in Beliaghata, a house that Manubehn Gandhi, who was with him, described as shabby, with broken doors and windows, one usable room and a single latrine for hundreds of people. He fasted while others celebrated. When he woke at 2 a.m., Muslims were already waiting outside; they had said they would break their own fast only after seeing him. The premier of Bengal, Huseyn Shaheed Suhrawardy, had agreed to live and work with him there, until every Hindu and Muslim in Calcutta could return home safely. Rioting broke out again on 1 September, and Gandhi began an indefinite fast, at the age of 78, and it ended only after written assurances from leaders on both sides. On the night of 31 August a crowd threw stones at the lamps and windows, and he answered, “Kill me, kill me.”
1948ChangchunIn 1948 the People’s Liberation Army ringed Changchun, in Manchuria, and from May to October kept civilians from leaving the city. A PLA lieutenant colonel, Zhang Zhenglu, wrote in a book published in August 1989 that about 150,000 civilians starved to death in the siege; that is one PLA officer’s account, in a book later suppressed. The Party’s own account called the battle nearly bloodless. An officer’s report dated 9 September 1948 says troops shot and beat residents who tried to flee. Zhang’s book sold more than 100,000 copies and was then suppressed in the spring of 1990.
1949BeijingThe phrase most people attach to the founding of the People’s Republic belongs to the earlier of two speeches. It was on 21 September 1949, to more than 600 delegates at the first plenary session of the People’s Political Consultative Conference, that Mao said the Chinese people had stood up. He named nationwide economic construction as the next task. The line is often placed on 1 October; Mao’s own text dates it to 21 September.
These scenes are not a rehearsal for the bread-or-freedom story. In Calcutta the first Independence Day was spent by Gandhi fasting among crowds who could confront him in person. At Changchun a besieging army kept civilians from leaving. China’s new rulers came to power through a victory that, for four decades, its official account described as nearly bloodless.
India’s founding vote, a ballot for every adult, is told in chapter 2.
What a state is founded to do, and what it can be told about itself, mattered more than whether it took a vote.
On 1 August 1958 the People’s Daily reported that rice production in a Hubei county had exceeded 7,500 kilograms per mu. The revised statistics put actual grain output in that province at closer to 120 kilograms per mu. The figure was published in the party’s own paper.
In 1959 the harvest fell by 15 percent. The state procured about 38 percent of what was grown, against 26 percent in 1958. Meng, Qian and Yared, economists who studied the famine’s geography, found that food production in 1959 was almost three times what the population needed to avoid starvation, and that the regions that produced more food per head suffered higher mortality, not lower. They attribute it to an inflexible procurement system that could not collect new information and act on it. Most of the deaths came in January and February of 1960.
In Xinyang, a prefecture of Henan, the local boss reported a 1959 harvest of 3.6 million tons when it was about 1 million. Officials confiscated 800,000 tons, the militia manned roadblocks so no one could flee, and roughly one person in eight died. In 1959 about 4.3 million tons of grain were exported to the Soviet Union. The Party’s own chronology records that on 1 January 1961 the Centre forwarded the Xinyang prefectural report to the whole Party, a report that blamed the trouble on bad people in power and on landlord and feudal sabotage, and it says this encouraged “Left” errors.
1962BeijingOn 30 January 1962, in front of more than 7,000 cadres, Mao said that any mistakes made by the Centre ought to be his direct responsibility. He added that many districts had not passed on his earlier self-criticism, “as if my mistakes could be hidden,” and that without a high degree of democracy there could be no high degree of centralism. The Party’s chronology records that the leaders’ disagreement over the main cause of the difficulties was not resolved.
How many died is disputed, and the book does not average the answers. Yu Xiguang gives 55 million; Dikötter more than 45; Yang Jisheng 36; Cao Shuji 32.5. Demographers reconstructing from census and vital statistics give 30 million (Banister), 23 million (Peng), 18.5 million (Yao) and 16.5 million (Coale). Chapter 2 gave Drèze and Sen’s narrower survey of the range. At the low end, Yang Songlin and Sun Jingxian calculate 3.5 to 4 million excess deaths, and Cheng Enfu and Zhan Zhihua 4 to 5 million unnatural ones. The spread is wide, and every figure on it is the death of a person whom the state decided whether to feed. The Party’s own 1981 verdict puts the cause chiefly on the Great Leap’s errors and the campaign against Right opportunism, joined by natural calamities and the Soviet Union’s scrapping of contracts.
The famine is the scene that matters for the argument about bread. Rule without a free press produced a famine in 1960 and a boom after 1978, so the form of rule alone does not predict the outcome.
Amartya Sen drew the sharpest version of the lesson. A famine of this kind does not happen in a democracy with a free press, he argued; without newspapers or an opposition, he wrote, the Chinese government came to believe its own propaganda, and thought it had a surplus of 100 million tons of grain. Drèze and Sen add that India has had no large-scale famine since independence. Democracy is superb at preventing visible catastrophe. It is much worse at forcing an investment whose payoff is invisible for years: a girls’ school, a clinic in a village that does not vote as a bloc, an industrial zone that will open after the minister has left office.
Put the two failures side by side. In one, a state could not be told the truth about a harvest, and so collected 38 percent of a crop that had fallen by 15 percent. In the other, a state cannot starve anyone without a newspaper saying so, yet it has had trouble building the things whose payoff comes late. Neither is a recommendation. Each is a failure with its own shape, one of information and one of choice, and only the first is recorded killing millions in these years. Chapter 2 told the basement China poured; chapter 7 told the step India sawed off.
For the first two of those slow investments, India has a witness on its own coast.
Kerala is a democracy, noisy and unionised. In 1957 its voters elected a Communist-led government. The Communist Party of India won 60 of 126 seats on 35.28 percent of the vote, while Congress won 43 seats on a higher 37.85 percent, and E. M. S. Namboodiripad became Chief Minister on 5 April. On 31 July 1959 the Union government dismissed it and imposed President’s rule, which lasted 206 days. Democracy is not tidy.
Its schooling is older than its democracy. In 1817 Rani Gouri Parvathi Bai, regent of Travancore, issued a royal rescript on education proclaiming that the state would bear the entire cost, the first formal recognition of education as part of the state’s administrative responsibility. Missionaries opened schools of their own in the same decades.
By 1951 Kerala’s literacy was 47.2 percent, against 18.3 percent for India as a whole. In the early-to-mid 1980s, 71 percent of Kerala’s adult women could read, against 56 percent in China (1985) and 26 percent in India, and Kerala’s life expectancy, about 66 in 1976 to 1980, was not materially different from China’s. Nobody had to be starved or silenced to get there. Drèze and Sen note that Kerala’s health and food programmes frequently followed articulated social and political demands, which is how a democracy is meant to carry a need from a village to a budget. By 2011 Kerala was 94.0 percent literate and India 73.0 (Economic Survey table; the age base for the 1951 figures differs).
Women who could read, early-to-mid 1980s
Kerala taught more of its women to read than China did
Adult female literacy, early-to-mid 1980s. Kerala's people also lived about as long as China's — and nobody had to be starved or silenced to get there.
Source: Drèze and Sen, Hunger and Public Action (Census of India 1981; UNICEF)
See the numbers
| Place | Women who could read |
|---|---|
| Kerala | 71% |
| China | 56% |
| India | 26% |
The job at the end of the schooling is another matter. In the first three months of 2024, 31.8 percent of Kerala’s urban residents aged 15 to 29 who wanted work did not have it, the highest rate of any state in the survey and nearly twice India’s urban figure of 17.0 percent. For young women the rate was 46.6 percent, against 22.7 percent across urban India. A school teaches a girl to read. It does not, by itself, give her somewhere to work.
That is the narrower charge the evidence supports, and it is not against democracy as such. It is of what India’s democracy chose to protect. Chapter 5 tells two of those choices, a car venture that had built almost nothing and a line above which an employer needed permission to let workers go. Each choice had a constituency that could vote. Nobody voted against a factory that was never built; its would-be workers were not yet a constituency.
The authors of Why Nations Fail (2012) would say the same story has a longer arc. In the book and in replies to critics, Acemoglu and Robinson argued that China’s growth under extractive political institutions would stall short of Spain’s or Portugal’s income per head. Separately, in 2012, Fukuyama argued that China is the hardest case for the theory, and that state capacity and rule of law matter more than democracy. Arvind Subramanian argued that China and India are outliers in opposite directions, and that the theory would need India to slide into authoritarian chaos, or China to turn near-Jeffersonian or collapse, to survive. Yuen Yuen Ang argues that the laureates’ picture of Western institutions is idealised, and that this is why the theory struggles to explain China. The Nobel committee gave the authors the 2024 economics prize for studies of how institutions are formed and affect prosperity. The question is still open. The book does not close it, and it asks something smaller: whether a democracy that knows what it protects can protect the slow things.
Chapter 8 left India holding its debt. The next chapter is the year the debt came due.