In 1781, the Qing Emperor executed 56 officials in the Gansu province for selling fake exam certificates. Their crime was not only one of corruption. In imperial China, the bureaucracy was seen as a fair one that promoted its members based on achievement. And for fifteen centuries, the foundation of that system had been the imperial exams. The Gansu officials had threatened the legitimacy of the imperial state itself—only the most severe of penalties would do.
Keju, the civil examination system, first appeared in the Han Dynasty in the second century BC and was maintained by nearly all dynasties since. Only brief interruptions took place under Kublai Khan and the Hongwu Emperor as they founded their respective new regimes. The Ming and Qing dynasties consolidated its modern form between the fourteenth and seventeenth centuries. Of the various forms of recruitment to the bureaucracy, Keju won out. The system was very selective. Only around one or two in a thousand attempting to become a shengyuan, or entry-level administrator, would eventually achieve the rank of jinshi, the highest official title.
The architects of Keju took pains to ensure the system was perceived as a test of real talent. A quota system determined the number of eligible candidates in each province, and one could take the exam as many times as one pleased. The imperial court jealously guarded the system’s reputation. While cheating remained an ongoing problem throughout its existence, the system never lost its reputation as an opportunity for advancement on the basis of talent.
Keju was designed to select for ability, not to achieve strictly equal representation across classes or redistribute access to state jobs. Moreover, those who made it into the system were able to pass on advantages in cultural formation and wealth to their children. Children in aspiring families began reading Confucian texts from the age of 6 or 7. Having parents who could afford to hire teachers and, later on, allow the child to forgo a job and instead prepare for the exam, certainly smoothed the road ahead.
The result was partial, but not absolute, advantages for the children of these families. Between a third and half of jinshi—as well as juren, the rank immediately below jinshi—came from common parentage. What was less obvious was that around 87% of juren came from families containing juren or jinshi within the previous five generations. Familial clans could and did develop strategies to maximize kin success in the exams by picking the smartest children in the extended family, directing funds for exam preparations to them, and having them take the exams consecutively until they succeeded.
While the success of individual families varied, kinship success in the exam remained relatively stable. Further, the quota system put no ceiling on the number of successful candidates from single lineages. Further down the line, a son who had succeeded in the exam could return to his hometown to pay back the debt, establishing education facilities, libraries, and other initiatives enhancing the prospects of new kin aspirants.
While largely hidden from plain sight, Keju influences from kin to kin created a reverse-persistent effect where parental background mattered less for social mobility than the culture of the entire kin network. Collaboration had to occur on the level of clans—large enough to regularly produce a couple of talented individuals per generation—rather than nuclear families.
Rather than abolishing familial advantage or inequality, Keju was a check against powerful families using their position to make up for incompetence. Under pressure from the Keju system, clans had an incentive to remain internally generative. A familial clan that could correctly identify and promote new talent from generation to generation could remain powerful for a long time. The system’s ability to retain top families meant that the social networks and collective knowledge of its members were also passed down generationally.
Although the emperors could never fully overcome the problem of cheating, families were incentivized to promote their best members and give them the proper training to pass exams and enter the imperial system. And to the degree that those from common backgrounds might demonstrate sheer talent, a higher station became a real possibility.
Benjamin Elman, the author of Civil Examinations and Meritocracy in Late Imperial China, explains how elites saw their success or failure:
Chinese used ‘fate’ (ming) to explain the social and cultural trends and inherent inequalities in the selection process. Many accepted their success or failure because they believed that the gods had determined the rankings beforehand. Elites, when unsuccessful in the examination competition, invoked fate to explain why others, who were not superior, succeeded.
The system endured until near the end of the Qing dynasty. As China’s inability to confront the West and deal with internal dissent became clear, reformers began to see the system as an outdated barrier to saving the country. Many of the late Qing reformers, such as the imperial councilor Ronglu, the general Li Hongzhang, and education reformers like Kang Youwei and Liang Qichao, had themselves taken the exams. The reformist faction believed that Keju’s focus on mastering old Confucian traditions no longer served society and dangerously misdirected elites. So in 1905, they made a last-ditch attempt to modernize China.
The End of Keju
The abolition of the exam system was abrupt. Japan had recently defeated Russia in the Russo-Japanese War. Chinese reformers attributed the victory to the modernizing effects brought about by the Meiji Restoration of 1868. Meanwhile, China had just experienced a failure to drive foreigners out of the homeland during the Boxer rebellion. On top of that, growth had been stagnant between 1895 and 1903, indicated by a decrease in the number of newly opened firms. 18 new firms had opened in 1894, but this number had shrunk to only three by 1895. The logic of the Chinese government was that if modernization had worked for Japan, perhaps it would for China as well.
In 1912, the first year after the fall of the empire, 125 new firms were established. Many Chinese students went abroad to Japan to acquire a Western-style education. Young adults with wealthy relatives could best afford to be privately tutored, travel abroad to Japan, or use their resources to establish a business. It would seem that not much had changed from the vantage point of social mobility.
In fact, the number of new firms or students abroad did not correlate that closely with financial wealth. What it did correlate with geographically was the regional quota levels of exam applicants during the Qing era. Regions with higher quotas of exam candidates saw more firms being established and sent more students abroad even after 1905.
Similarly, application durations to enter Japanese schools also decreased in proportion to quotas. After the end of Keju, those who had sought advancement via the exam system were now doing so by building businesses and receiving a modern education. However, social mobility remained almost as low as during the Qing period. The position of the Qing elite continued to give them advantages in the republic period. They were still overrepresented in official positions as well.
But despite these advantages, the environment had changed. Many of China’s old elites sensed that new skill sets were needed in this era. Despite low social mobility overall, cultural and economic changes still made their individual positions uncertain. During the Qing era, ambitious Han Chinese—who the Manchu Qing dynasty placed under numerous social restrictions—often sought advancement through the imperial exam system. In its decline, many turned to revolutionary activity instead. Higher imperial exam quotas in a region correlated positively with revolutionary participation after the abolition of the exam system. Aspiring Han Chinese contributed to the 1911 republican revolution and to the final end of imperial China.
The fact that modernization did not greatly impact social mobility shows that the Qing reformers had been correct in their assessment. The problem with Keju was not that it failed to cultivate talent properly, gatekept for incompetent elites, or suppressed mobility. In the absence of bureaucratic jobs, many Keju-selected Qing-era elites had still proven their abilities. They embraced modern education, succeeded in the new business landscape, and operated in republican institutions.
Rather, Keju had allocated talent in the wrong direction, at least in terms of economic growth. Until 1905, it had kept China’s talent from pursuing modernizing activities by incentivizing them to largely maintain the elite culture as it had existed for centuries.
When Mao Zedong became the leader and founding father of the People’s Republic of China in 1949, the success of the imperial elites made republican modernization a target of the regime. Throughout Mao’s reign, including during the Cultural Revolution, he attempted a complete redistribution of wealth and power. Land reforms expropriated property from landlords, schools were closed, and values such as traditional learning and social status became abhorred.
Mao succeeded, initially. For the generation that grew up during the Cultural Revolution, the educational gap disappeared. Old inequality patterns not only straightened out, but reversed across counties—but not by enhancing educational prospects at the bottom. Instead, the Maoist system removed educational prospects at the top. In other words, the revolution achieved equality at the expense of creating educational and material wealth.
Desperate elites might attempt to hide assets and wealth—for example, by slaughtering and consuming cattle before they turned into collective goods—but such activities were unlikely to be more than ephemeral gusts of resistance. Such equalization ultimately came at the cost of sustained Chinese growth, which waxed and waned at the mercy of various political campaigns, and through a political culture that spent most of its energy on round after round of purges.
After Mao’s death, the Gang of Four’s purge, and the rise of Deng Xiaoping, China’s leadership reordered its priorities. From 1978 onwards, growth was back on the agenda.
The cards were stacked against the re-emergence of old mobility patterns. The descendants of China’s imperial elite were left without education, wealth, or property. The parents of children born during the reform era had generally not attended secondary school or university. Any wealth had been confiscated by Maoist officials, leaving no room for its transmission from grandparents to parents. Further, whatever land had been expropriated was not given back to its previous owners during the reform era. Rather, it was reallocated through lotteries or auctions, a decollectivization controlled by county officials rather than village cadres in order to minimize possible corruption.
Despite this near-total equalization in their material starting points, a familiar trend slowly became evident as the era of Dengist reform went on. China’s pre-revolutionary elites were making a comeback.
The effect was not just limited to a few descendants of high officials, but to the broader class of landed gentry that had long been demonized as the source of China’s decline. Studies done on the post-1978 generation show that a child whose grandfather was a rich landlord before the PRC, but whose parents had no physical capital to pass on, still tended to fare better than their materially equal counterpart with no such heritage.
In 2010, those whose grandparents had been part of the landowning pre-revolutionary elite earned 16-17% more each year than those whose grandparents were not. They have an unusually high rate of persistence today too, with a 14.5% chance of staying in the top 10% of income-earners for three generations—a higher rate of persistence than their counterparts in the U.S., Russia, and Taiwan.
If beginning from lowly occupations, China’s elites have far higher chances of advancing. For those whose parents worked in agricultural jobs, the children of pre-revolutionary elites have a 33% higher chance of advancing to a non-agricultural job. And even the long-dead Keju continued to shape Deng’s China: in 2010, those living in an area with double the density of jinshi had, on average, 8.7% more years of schooling. For those who had grandparents among China’s landowning elites, the number was 11% more than non-elite peers.
China’s elites were making their comeback without any inherited material advantage. Mao had managed to shut down channels of persistence related to inheriting wealth and institutional power. But he could not stamp out the transmission of educational mores, culture within the household, or genetically inherited talent.
As revolution and reform raged about them, such families continued to form their children within the household. It was a trait they passed down: the descendants of China’s landowning elites are still far more likely to live with their parents and grandparents, who in turn are more likely to invest in education for their children. Even today, the children of former landowning families continue to outpace their non-elite peers in reading.
Elite families also maintained a different attitude to work from their peers. Surveys on work ethic found that they were far more likely to see hard work as important for success and kept doing so as they advanced. Children watched their parents work significantly more hours than non-elite peers and maintained the same ethos.
The Maoist regime also could not erase the deep ties between elite families. During reform and opening up, these relationships became assets as well. Today, in counties where the concentration of elite surnames is higher by one standard deviation, the income gap between elite and non-elite is higher by around 31%.
The advantage maintained by these families wasn’t extractive institutional privilege, but their ability to pass down and develop positive-sum cultural wealth. As China embraced economic growth, these mores proved to be powerful assets for elite families entering the new industrial market economy and for the growth of the economy overall.
Persistence Everywhere
Looking across the globe, the Chinese pattern of persistence does not look like a freak accident. Many societies that were completely disrupted by revolutionaries with radical redistributive policies in the twentieth century had similar experiences. Egalitarian revolutions around the world fared just as poorly as China in achieving their long-term re-distributive goals, only achieving the destruction of cultural and material wealth. Some rocks, it turns out, survive earthquake after earthquake.
The socialist experiment of the Hungarian People’s Republic had a far more liberal policy toward consumer goods and political freedom than Mao’s. Yet, it yielded similar results to the Chinese one when it came to elite persistence. In 2010, if you happened to be a descendant of the aristocracy from eighteenth-century Hungary, you were 2.5 times as likely as the average population to gain a medical qualification.
A similar pattern unfolded in America. Tens of millions of people arrived from Europe to the “land of opportunity” in the years from 1850 to 1913. However, 51% of the gaps in occupational income present when the first generation arrived were also present three generations later.
Glancing at these results, one might be inclined to believe, as the economist Vilfredo Pareto did, that “in all places and at all times, the distribution of income remains the same,” a situation he considered a law of social science. Perhaps success is purely hereditarian, with education and culture being fairly pointless expressions of inherent ability.
But as a case study in persistence, China’s elites show that culture is irreplaceable. Individuals, however talented, do not advance on their own. Developing talents is a complex process that involves parental example, formal and informal education, a familial network that sets examples and creates opportunities, and transmitting tacit knowledge that is hard to learn theoretically. Moreover, nuclear families often regress to the mean, making parental heritage useless on its own. Only extended kin networks with a culture of rigorously identifying and cultivating talent can endure more than a generation or two.
But contrary to Maoist assumptions, the fall and rise of China’s elites did not come at the cost of Chinese development. On December 13th, 1978, the newly inaugurated Deng Xiaoping addressed the nation in a speech titled Emancipate the Mind, Seek Truth From Facts, and Unite As One In Looking to the Future. Deng was adamant that China had been held back for too long by rigid adherence to an inadequate philosophy, and that things had to change. But he did not set out to change social mobility. He set out to generate wealth.
Many comrades have not yet set their brains going… their ideas remain rigid or partly so… During the past dozen years Lin Biao and the Gang of Four set up ideological taboos or “forbidden zones’” and preached blind faith to confine people’s minds within the framework of their phony Marxism. No one was allowed to go beyond the limits they prescribed; anyone who did was tracked down, stigmatized and attacked politically… some people found it safer to stop using their heads and thinking questions over… [and] once people’s thinking becomes rigid, [then] book worship, divorced from reality, becomes a grave malady.
Deng’s speech did incite a tremendous amount of activity in the Chinese population. The poverty rate fell from 97.5% in 1978 to 3.1% in 2017.
Elite persistence and national development had never been mutually exclusive goals. As the Qing reformers had realized, elite persistence had never been proof of the Keju system’s dysfunction. Rather, it was a problem precisely because of its functionality. By devoting resources to cultivating talent within the old imperial system, China’s elites were reinforcing a barrier to China’s economic development. It was the system that had to go, not the elites.
As wealth stagnates and conflicts grow in our own society, it can be tempting to see elite persistence as one more political failure. But the lens of the Qing reformers offers a different read: in a system held back by stagnant institutions and obsolete cultural assumptions, the future doesn’t depend on elite circulation or redistribution as such. Instead, the question is whether elites, old or new, can anchor their success in a new and more functional order before it’s too late.