I can't claim to have a broad enough perspective on Chinese history to be able to describe the changes at length; that said I can discuss a couple of approaches to this question that are worth going into.

Firstly, state structures in China absolutely did change over time. A paradox of the 'dynastic cycle' model is that it can imply that new dynasty-states represented decisive breaks in continuity that were not always the case, or that dynastic change merely meant different ruling houses coming into the possession of a fundamentally unchanging underlying structure. '2000 years of imperial rule' is a fun shorthand but it can lead to assuming that imperial rule in 200 BCE was the same as in 900 CE as in 1800, which, well, it wasn't.

One of the primary alternatives to the dynastic schema for periodising Chinese history is that proposed by Jacques Gernet, based on changes in Chinese political forms and transcending the bounds imposed by individual dynasty-states. As someone who cannot claim even the remotest expertise pre-1600 I cannot say with any firm certainty how far his scheme is considered accurate in present-day academia, but it nevertheless serves to illustrate that there were visible political changes:

Time	Gernet's Classification	Dynasties
1600-900 BCE	Palace Civilisation	Shang
900-500 BCE	Autocratic Cities	Shang, Zhou
500-220 BCE	Development of Monarchical Institutions	Zhou
220 BCE-190 CE	Conquest of Former Kingdoms	Qin, Han
190-310 CE	Military Warlords	Han, 'Northern and Southern' Period
310-590	Military Aristocracy	'Northern and Southern' Period, Sui
590-755	Sino-Barbarian Autocracy	Sui, Tang
755-960	Military Adventurers, Division	Tang, Five Dynasties and Ten Kingdoms
960-1280	Reunification	Song, Liao, Jin
1280-1370	Non-Chinese Empire	Yuan
1370-1520	Autocracy	Ming
1520-1650	Political Crisis	Ming
1650-1800	Peace and Prosperity	Qing
1800-1900	Collapse and Loss of National Independence	Qing
1900-1950	Military Dictatorship, Peasant Militias, Founding of PRC	Qing, Republic of China, People's Republic of China

Now, Gernet's model definitely has issues, particularly in its presentation of the Song as maintaining territorial integrity (much of northern China would be lost to the Jurchen Jin) and in its assertions of Qing continuity from the Ming. But as noted, there are clearly ways to approach Chinese political and institutional history that highlight change rather than continuity.

For my own part, looking at the Qing, it is true that superficially, the Qing retained many of the structures of the Ming state intact, within China proper anyway. But the wider organisation of the Qing Empire was largely unique. For instance, the Eight Banners formed a distinct caste of trusted soldiers, officers, and administrators, based on lineage and/or ethnic lines, distinct from anything that had existed under the Ming. A degree of 'ethnic sovereignty' also prevailed, and Manchus and Banner-enlisted Mongols were particularly preferred for promotion in civil office, leading to Manchus being disproportionately overrepresented at high levels of government: over the course of the Qing period, 48% of provincial governors and 57% of viceroys (a position above that of governor, created by the Ming and expanded in scope by the Qing) were from the Banners, not the regular 'civilian' bureaucracy. Of course, on top of that, those gubernatorial positions themselves gained substantial prestige and importance relative to under the Ming, with a much firmer delineation of roles and responsibilities.

If we want, we can assess the Qing against all of the features you've delineated in your post:

highly centralised

Is technically true insofar as we mean that much power was held by the state, though we ought also to understand how relatively decentralised normal decision-making could be. In effect, the emperor rarely issued much legislation purely on personal initiative; rather, officials at the local and provincial level were expected to make decisions on their own initiative where possible, reporting up the chain to people with veto power over those decisions. The emperor, then, often functioned reactively to his governors' activities.

uber-efficient bureaucracies

Efficiency will always be a relative matter, but the proportion of civil servants in the Qing was never that high. There were a total of 1500 county-level magistrates, the lowest level of the provincial hierarchy; in 1800 there were perhaps 20,000 total official posts. At a time when the population numbered some 400 million, this meant that there was one official of any sort per 20,000 people, and each county magistrate was, on average, responsible for nearly 300,000 constituents. Both by necessity and to an extent by design, Qing government at the local level was deeply reliant on cooperation with local elites with entrenched local interests who provided assistance of various sorts to officialdom in exchange for both explicit favours and those officials' recognition of local issues.

systems of highly competitive state-run schools that churned out armies of very educated mid-level administrators

As noted, the Banners are a weird dimension of this in that Banner officials were not necessarily classically educated the same way as most civilian officials were. We should not exaggerate the extent to which they are supposed to have lacked the level of education that other officials had, but the nature of that education could be somewhat different. Indeed, a number of Banner officials had been trained as translators rather than as classical scholars. Office-selling was also common – in the 1760s, some 20% of low-level official postings had been obtained by purchase rather than assigned through examination success.

strict laws that were universal across provinces, which themselves rarely exercised any real regional autonomy

That is very much untrue: the upper provincial officials' relationship with the emperor was a complex one. On the one hand they were a select, trusted cadre: what Philip Kuhn terms the 'provincial bureaucracy' in the narrow sense consisted of less than 100 people: 8 viceroys, 17 provincial governors, 18 provincial treasurers, 18 provincial judges, and a handful of trans-provincial commissioners like the superintendents of the Grand Canal and Yellow River. On the other, that level of closeness to the emperor is also what allowed them to bend the rules a little and expend some of their social and political capital towards their own ends. The provinces themselves were not particularly autonomous, in the sense that few elites had province-level interests, and the 'law of avoidance' preventing officials serving in their home provinces, along with regular rotations, prevented the buildup of local power bases. But the officials running those provinces could and did exercise a considerable degree of initiative at times.

Moreover, if we look beyond China proper we see a much more complex system of compromises and impositions. The Qing ruled over Manchuria, Tibet, Mongolia, and the Tarim Basin, all of which meant reckoning with a wide variety of administrative systems and political cultures and which entailed a much more heterogeneous approach to governance across the wider empire. Mongolia had the jasak chiefs, Tibet retained its Mongol-founded government, the Ganden Podrang, and the Tarim Basin's city states were largely delegated to local officials known as haqim begs. There were also variations in China proper because of substantial areas of autonomous indigenous territory in Yunnan, Guizhou, and Taiwan, which meant that there were a few provinces where the Qing administrative presence was basically like Swiss cheese, with large pockets of free indigenous territory, at least until, during the early eighteenth and late nineteenth centuries, these were eroded and eventually erased in a series of colonial projects.

Conclusions

While the Qing never represented a complete civilisational break from the Ming, there were nevertheless several noticeable areas of discontinuity from their Han Chinese predecessor, even without taking the wider empire into account. More importantly, the idealised hyper-bureaucratised model of Chinese statehood simply does not fit the Qing, whose officialdom was never a homogeneous extension of the imperial will, but a highly variable structure capable of having its own agenda.

Suggested Reading

• Philip A. Kuhn, Soulstealers: The Chinese Sorcery Scare of 1768 (1990) – this goes into great depth on not just the social issues involved, but more importantly how the Qing state responded to the scare, with a particular spotlight placed on tensions between the emperor and officialdom during the crisis.

• R. Kent Guy, Qing Governors and Their Provinces: The Evolution of Territorial Administration in China, 1644-1796 (2010) – A great overview of the provincial system in the early and middle Qing period, with a detailed breakdown of individual provinces' issues.

• Wensheng Wang, White Lotus Rebels and South China Pirates: Crisis and Reform in the Qing Empire (2014) – A slightly misleading title, but it goes into depth in the institutional crisis faced by the Qing amid the White Lotus Revolt, and the administrative reforms that went into effect as a result.

• Bradly W. Reed, Talons and Teeth: County Clerks and Runners in the Qing Dynasty (2000) – Discusses the bottom level of the Qing administration that lay outside the examination hierarchy.

• Peter C. Perdue, China Marches West: The Qing Conquest of Central Eurasia (2005) – Contains some discussion of Qing administration in Inner Asia.

• Mark C. Elliott, The Manchu Way: The Eight Banners and Ethnic Identity in Late Imperial China (2001) – The landmark work on Manchu identity under the Qing, useful framing in general.

The case of Wang Mang and his (short-lived) reforms was the first to come to my mind, and this was also brought up earlier here by u/mikedash.

The official selection process of civil service has evolved considerably in Imperial China, and it can definitely be one example of a governance and bureaucratic change with time.

During the Han dynasty, a system of recommendation of talents by officials was established as a means to select officials for positions. In 165 BCE, Emperor Wen of Han decreed to have his officials and royal relatives to recommend people who have outstanding talents, good conduct, and who can speak up with integrity to him, in order to correct his oversights^(1). In 134BC, Emperor Wu of Han decreed to have commanderies recommend people of talent and virtue^(2). In 106BC, when Emperor Wu established the Inspector system, he also ordered those high ranking officials in respective provinces to recommend people of extraordinary talent (茂材; Mao-cai) to become officials^(3) .

Emperor Wu also established the Taixue, or the Imperial Academy, with Erudites appointed to teach the 5 classics^(4). People, either through recommendation or selection by higher officials, can become students of these Erudites^(5). These students can then participate in tests known as 射策 (She-ce), in which they will randomly choose one particular question and answer them. They will then be graded based on difficulty and performance, and given official positions afterwards. A number of officials were recorded to have obtained their positions through this process, such as Xiao Wangzhi, who eventually became the Chamberlain for Attendants, a considerably high-ranking title^(6).

This system was one principally based on recommendation, and soon issues arose.

Since recommendations were very important, many people began to purposely build up their reputation so that they can get noticed and recommended by local officials. There was a story of how an elder brother purposely took a lions’ share of the inheritance such that his younger brothers would be praised for their virtue (not contesting with the elder brother) and get recommended^(7).Many local powerful families also used their local influence and power to ensure that they get the recommendations from local officials, in one particular case 5 out of 6 recommended were due to their wealth and connections^(8).One official also commented that officials like to recommend young people because they can often repay the favour to express their gratitude, while the older ones might not have the time to do so^(9). An Easter Han nursery rhyme was also recorded to have mocked how those who were recommended as talents can’t even read, while those who were recommended for being virtuous were in fact uncaring towards their parents^(10).

In was in this context that some reforms and changes were proposed. In 132CE, Easter Han’s then Director of the Imperial Secretariat Zuo Xiong proposed a revamp of the system. He proposed that those who were recommended must be above the age of 40, and mandated tests for all the recommended individuals before they can become an official^(11). Those who recommended people who were deemed untalented later were also heavily punished. As a result, officials were afraid of the punishments and dare not recommend out of free will without considerations^(12).

As we can already see from this very short and summarized glimpse, the official selection process itself has changed considerably even just within the broad Han Empire (albeit Western and Eastern to be more precise). And when we compare this with even later selection processes centuries later– The Imperial Examination system that Imperia China is famous for – we can notice even bigger changes.

In later periods where such civil examinations were more established, we start seeing more refined and streamlined segregations of grades and regions, with elaborate names like Presented Scholar (Jin-shi 進士) and tiers going progressively from the local town to the palace examinations.

Participation had also changed considerably. The Han’s selection process were heavily dependent on the basis of recommendation. Even for the tests under Zuo Xiong, only people recommended were eligible to go through such tests to prove their capabilities. Whereas by the Tang dynasty, people could self-register for civil exams in their respective county and province^(13). Unlike the Han’s selection tests under Zuo Xiong in which all participants who met the requirements will theoretically be given positions, later imperial examinations have participants numbers far greater than the positions available; and one not only have to pass the requirements, but also do better than their peers in order to gain a position. This thread on How hard was the Chinese Imperial Examination touched on some of that pretty nicely.

References

• The Cambridge History of China: The Ch'in and Han Empires
• Chinese History: A New Manual
• 察举制度变迁史稿
• 漢代察舉制度考
• Relevant sources cited in footnotes

​

1. 《漢書·文帝紀》:十五年…..九月，詔諸侯王公卿郡守舉賢良能直言極諫者
2. 《汉书·武帝纪》:元光元年冬十一月，初令郡國舉孝廉各一人。
3. 《汉书·武帝纪》：“元封五年，初置刺史部十三州。名臣文武欲尽，昭曰：‘盖有非常之功，必待非常之人。故马或奔踶而致千里，士或有负俗之累而立功名，夫泛驾之马，跅弛之士，亦在御之而已，其令州郡察吏民有茂材异等，可为将相及绝国者。’
4. 《汉书·武帝纪》:置五經博士。
5. 《漢書·儒林傳》: 為博士官置弟子五十人，復其身。太常擇民年十八以上儀狀端正者，補博士弟子。郡國縣官有好文學，敬長上，肅政教，順鄉里，出入不悖，所聞，令相長丞上屬所二千石。二千石謹察可者，常與計偕，詣太常，得受業如弟子。
6. 《漢書·蕭望之傳》:望之以射策甲科為郎…..及宣帝寢疾，…..望之為前將軍光祿勳，
7. 《文獻通考·選舉考七》:許荊祖父武，太守第五倫舉為孝廉，武以二弟晏、普未顯，欲令成名，乃請之曰："禮有分異之義，家有別居之道。"於是共割財產，以為三分，武自取肥田、廣宅、奴婢強者，二弟所得悉劣少。鄉人皆稱弟克讓，而鄙武貪婪，晏等以此並得選舉。武乃會宗親，泣曰："吾為兄不肖，盜聲竊位，二弟年長，未豫榮祿，所以求得分財，自取大譏。今理產所增，三倍於前，悉以推二弟，一無所留。"
8. 《後漢書·張王种陳列傳》:今當舉六孝廉，多得貴戚書命，不宜相違，欲自用一名士以報國家，爾助我求之。
9. 《後漢書·樊宏陰識列傳》:上言郡國舉孝廉，率取年少能報恩者，耆宿大賢多見廢棄
10. 《抱朴子》:靈獻之世，閹官用事，群奸秉權，危害忠良。臺閣失選用於上，州郡輕貢舉於下。夫選用失於上，則牧守非其人矣；貢舉輕於下，則秀孝不得賢矣。故時人語曰：「舉秀才，不知書；察孝廉，父別居。寒素清白濁如泥，高第良將怯如雞。」
11. 《後漢書·左周黃列傳》:雄又上言：「郡國孝廉，古之貢士，出則宰民，宣協風教。若其面牆，則無所施用。孔子曰『四十不惑』，禮稱『強仕』。請自今孝廉年不滿四十，不得察舉，皆先詣公府，諸生試家法，文吏課牋奏，副之端門，練其虛實，以觀異能，以美風俗。有不承科令者，正其罪法。若有茂才異行，自可不拘年齒。」帝從之，於是班下郡國。
12. 《後漢書·左周黃列傳》:明年，有廣陵孝廉徐淑，年未及舉，臺郎疑而詰之。對曰：「詔書曰『有如顏回、子奇，不拘年齒』，是故本郡以臣充選。」郎不能屈。雄詰之曰：「昔顏回聞一知十，孝廉聞一知幾邪？」淑無以對，乃譴卻郡。於是濟陰太守胡廣等十餘人皆坐謬舉免黜，唯汝南陳蕃、潁川李膺、下邳陳球等三十餘人得拜郎中。自是牧守畏慄，莫敢輕舉。迄于永嘉，察選清平，多得其人。
13. 《新唐書·選舉志上》:而舉選不繇館、學者，謂之鄉貢，皆懷牒自列於州、縣。

Though dynasties rose and fell, the way they rose was not always the same and the way they rose would impact what power they had at the centre, what legitimacy they might call upon, what challenges they faced. It is very easy to see declines as the pattern of loss of moral authority, corruption, ineptitude but dynasties fell for varying reasons that reflected the problems of their times.

A dynasty itself could change. I don't mean just West Han and Eastern Han dynasties were different: the three excellencies rather than a chancellor leading the government (the three kingdoms civil war saw regimes split on that), reporting officers rather than province heads going to capital each year, end of conscription (bar border areas) with a reliance on a small professional army to supplement local forces when required, iron and salt monopolies to the local authorities rather than central control, an attempt to be more austere. I mean in the sense that Eastern Han that started that unified the land in 25 CE would be different from the one that would be struggling in the 180s before it collapsed.

Things changed over decades, centuries, rulers, events. Each Emperor with power, each Dowager, each regent via General-in-chief, even the more active political players at court could have an impact. Even a straight father to son (rare for the Eastern Han) succession meant a change of personnel at the top, someone who might hold different attitudes and different experiences from their father. The life experience of Emperor Ling who grew up in poverty, the start of reign saw an attempted coup and would not leave the capital on becoming Emperor was going to shape how he ruled. The Wei Emperor Cao Rui's attitude towards the new wave of scholars like He Yan, Xiahou Xuan and Zhuge Dan, backing the conservative complaints and seeking ways to restrict them from office, would be different when Rui died and the regent Cao Shuang embraced men like He Yan and Xiahou Xuan before his destruction by Sima Yi and the conservatives in 249, partly for embracing the neo-Daoists

To use Shu-Han, which had just two Emperors, as an example: Liu Bei who founded it would grow up having wave grass-mats and sell sandals due to family poverty, had an education from one of the best teachers due to a relative, became a fighting man who travelled across the country, highly ambitious, charisma, generosity, kindness resilience, experienced command, political skill (including willingness to backstab) and mastery of reputation. His son Liu Shan seemingly had a disrupted education, was kidnapped twice but of rather bigger status in his youth as the son of a warlord. He shared the kind spirit of his father but was lacking in brains or energy, he loved to travel to see the sights and tended to leave power among those he trusted. His regime and its shifts would be shaped by those who held his trust but also by Liu Shan.

However, this risks putting a bit too much emphasis on the Emperors and the powerful. They mattered of course but they and dynasties were also shaped by things that changed around them that they didn't have control over. For the Eastern Han, there were challenges from abroad (not counting the waves of epidemics from the 160s onwards), from changing philosophical and religious beliefs, the strain of a broken tax system, the movement of the populace from the north to the south leaving northern frontiers undermanned. What to do about an Imperial University that became increasingly irrelevant and untrusted politically and in scholarship? What to do about the powerful local families who, over time, turned away from the Han, had their own armed retainers and were dangerously powerful?

Each challenge was something a dynasty and the people of its time had to respond to. Sometimes the changes of policy were effective and stuck, other times it would last however long the figure at the top remained there, sometimes things were not fixed (like recruitment, like the university) despite attempts from a ruler or his regent. These challenges and the attempts to meet them would shape the dynasty and what would happen after.

A time traveller from early in the Eastern Han to the last decades would find so many things they might not entirely recognize from their own days: A financially strapped dynasty using fines for office in a way never intended, a new school to reflect new literary arts in calligraphy and rhapsody, Emperors and court officials with an interest beyond the traditional texts and beliefs (as well as people on the ground with faith-healing on the rise as a response to the epidemics) while old text vs new text remained. Emperors who now barely saw outside of Luoyang since taking the throne, a University so mistrusted with academic texts that the Stone Classics had to be created so there was a reliable version of the Classics, even marriage away from the traditional seven families.

A corrupted recruitment process with a rise in patronage and clients, the rise of refusal to serve as a way of showing your morality, vengeance killings with the backings of political thinkers of their day. The rise of eunuchs as an important and powerful faction against the gentry after repeated rescuing of Emperors from over-mighty subjects, the change of military policy during the great Liang rebellion of 107-117 to use a better system of fortifications, smaller but better-equipped forces and more use of cavalry for mobility.

There were changes in the bureaucracy, the role of General-in-Chief changing and changing again with the Liang's becoming chief ministers under Emperor Shun and the role's the connections to the military. Who had control of the Secretariat changed, away from Emperors towards leading officials and the General-in-Chief's. The use of Imperial Clerks as agents of the Emperor included leading armies in times of crisis to oversee forces that went beyond the provincial level, the brief introductions of exams for those nominated to office. The frontier provinces went from the surveillance system of Inspectors to the more powerful Governors to try to deal with the crises there following the Turban revolt. The Excellencies no longer always serving in the capital as the likes of Zhang Wen were needed on the borders.

An existing dynasty is not static, fixed in time to the starting point. Each generation adapted to the changing times and challenges, factions be they eunuch, reformers, in-laws, major gentry factions, those with an interest in central authority and those of the regional figures who wished for lower costs and were not so keen on expansion.

A new dynasty didn't tend to, at least in my area of coverage, go "same again please." The last dynasty fell and there would be reasons for that, the new wave of Emperors would not be seeking to go through the exact same thing. The new dynasty would seek to learn from the mistakes of the past, creating policies to try to avoid repeating the same patterns. They would also have to be coming up with answers to the challenges of their time which would not always be the same as the last dynasty.

After decades of civil war, in 220 the last Eastern Han Emperor Xian abdicated to the Cao family of the Wei dynasty with Cao Pi as the first Emperor. The Cao's thanks to Cao Cao had risen from a minor warlord and gentry family into the dominant figures in the civil war, controlling the central plains as well as the Han Emperor while establishing a court at Ye of intellectual quality with the Cao family having many a poet. Cao Pi had succeeded Cao Cao as King and, though the land was not unified, became Emperor.

The Cao's sought to fix the mistakes of the Han. The agricultural colonies of Ren Jun and Zao Zhi had been a key military move in Cao Cao's rise but it also tied people and resources to the central authority, countering the farms that had become under increasing regional family control. There was a belief that the Han had declined partly due to lack of enforcement of the laws, that the powerful had been treated too leniently, that amnesties had become a damagingly regular part of life. So the Cao's sought to restore central authority with, usually, a strict regime as well as one that had an intellectual flourishing heart. They married lower down than the Han Emperors had and sought to restrict the power of the imperial in-laws, regents were not going to be Dowager and clan but selected officials with half being from the imperial clan. There were attempts to fix the recruitment system that had fallen into the hands of local powers during the Han but Cao Cao's attempts to widen it were stymied by officials like Cui Yan and Mao Jie while Chen Qun's grand reforms of the nine ranks would go wrong with the rectifiers overstretched and easily in the hands of the powerful.

When Cao Shuang's regime, at the intellectual height of Wei, sought to further centralize power and fix the recruitment system, it lost effective power with a coup led by fellow regent Sima Yi. Cao Shuang and his allies were accused of treason and debauchery, executed. The Sima family, despite plots from figures at court and military revolts by several leading generals, would ruthlessly hang onto power. Even resorting to deposing one Emperor and killing another in the streets of the capital. After son Sima Zhao was able to get Shu-Han destroyed in 264, grandson Sima Yan would found the Jin dynasty in 266 with Cao Huan abdicating and in 284, Jin would unify the land.