Explaining British Political Stability After 1832

Author(s): Davis, Donagh; Feeney, Kevin C. | Abstract: Though not its main focus, Goldstone's Revolution and Rebellion in the Early Modern World (1991) threw considerable new light on 19th century Europe's revolutions and near-revolutions. While Goldstone stresses the role of an expanding and industrializing economy in absorbing 19th century England's demographic shocks, we accept this analysis but argue alongside it for similar attention to the vector of emigration, settler-colonialism, and imperial state expansion into which at least some of the exhaust fumes of the population explosion were vented. Furthermore, it is important to note the crucial role of a highly interventionist state and 'big' government in the background to these dynamics—a far cry from the light-touch, laissez-faire qualities with which the 19th century British state is often associated.To make our case, this article takes advantage of secondary literature and raw data not available prior to the publication of Goldstone's book. Of crucial importance here is our unique dataset of fatality-inducing political violence events in Britain and Ireland from 1785 to 1900. This is the first research paper to utilise this dataset. We draw upon this in the following section, which seeks to establish what the real level of political instability was in 19th century Britain—thus cross-referencing Goldstone's account with more recent data—before moving on in the following section to a more detailed overview of the socio-economic conditions underlying events at the political level. This is followed by our account of the emigration-settler-colonialism-imperial state expansion vector and the interventionist state policy behind it, which we argue was crucial to making 19th century Britain relatively 'revolution-proof'—alongside the expanding economic opportunities rightly highlighted by Goldstone. Lastly come our brief concluding remarks, which lay out the implications, as we see them, of this article's findings for research on revolutions, political violence and instability, demographic-structural theory, state-building, migration, and imperialism-colonialism.


Introduction
Though not its main focus, Goldstone's Revolution and Rebellion in the Early Modern World (1991) threw considerable new light on 19 th century Europe's revolutions and near-revolutions.Goldstone argued against several familiar explanations of these events-from the Marxian view that they reflected the new class tensions of the industrial age, to the idea that they were not 'real' revolutions at all, but political epiphenomena without deep social-structural roots. 1 Rather than reflecting the 'newness' vaunted by Marxists, or petty political squabbles, Goldstone contended that they were in reality the last gasps of Europe's age-old pattern of periodical, demographically-influenced state crisisalbeit exacerbated by the quickening pace of demographic change in this era.Furthermore, Goldstone compared England's 19 th century experience to that of continental European polities (particularly those of France and pre-unification Germany) and asked why England diverged so much after 1830 or so-a pertinent question, given Goldstone's emphasis on the transnational nature of the structural tensions tending to underlie state crises and revolutions (and explaining why they often come in waves).Goldstone tracked similar political stress indicator (psi) trajectories up to 1830 for France and England, before England's psi curve tapered off toward relative stability afterwards, while France and other continental states saw further violent uptick in psi in the ensuing years.Goldstone's analysis extends as far as the period around 1848-a tumultuous year of attempted revolutions across continental Europe, but of relative stability in England.However, it is worth bearing in mind that if this analysis were extended to the later 19 th century, it would have to reckon with further bouts of serious social unrest in France and continental Europe-notably the 1871 Paris Commune-and continuing relative stability in England.
Goldstone explained this divergence mainly in terms of England's far superior capacity to absorb increasing demographic pressure through its rapidly expanding, and rapidly industrializing, economic infrastructure, with its burgeoning appetite for labour.Twenty-five years on, this article looks again at this historical vista, and revisits the question of Britain's divergence after 1830 from the wider psi patterns seen in France and elsewhere in Europe.While accepting virtually all of Goldstone's groundbreaking structural-demographic analysis of these cases, it asks whether this analysis can entirely explain England's or Britain's notable divergence-or, in other words, explain why there was no revolution or significant revolutionary situation in 19 th century England or Britain comparable to those of her European neighbours.The article goes on to suggest that while Goldstone's insights are crucial to explaining what happened (or indeed did not happen), they may nevertheless leave some significant gaps unfilled.Indeed, while Goldstone portrays England's demographic-structural pressures as becoming less severe than those of France as the century wore on, viewed from another angle it is possible to see in this comparison a rather different story: one in which it was Britain, not France (or Germany, etc.), experiencing by far the most severe demographic-structural pressures at this time-making Britain's escape by 1830 from the cycle of revolutions and nearrevolutions all the more remarkable.For while France's population grew by 31% between 1800 and 1850, and Germany's by 61% between 1816 and 1864, England's grew by an astounding 92% between 1800 and 1840 (Goldstone 1991: 290).Indeed, between 1750 and 1850, it almost tripled (Goldstone 1991: 323).
This was an unprecedented demographic explosion, for which it is difficult to find meaningful comparisons before or since-and this article argues that England's expanding industrial economy, extremely dynamic though it was, was necessary but not sufficient on its own to absorb these new demographic pressures.We argue that a more complete answer to our question must take account of certain other fields in which 19 th century England and Britain were also a world leader-and we have in mind here the intricately intertwined histories of emigration, settler-colonialism, and imperial state expansion (ESCISE).With the partial exception of Ireland (at this time fully part of the UK), England exported much more of her population-both in absolute and per capita terms-than any other country in 19 th century Europe.Although the concept of an English (or for that matter British) 'diaspora' has never been en vogue-in so far as it registers as a concept at all-the 19 th century English (or English and Welsh, or British, depending on one's perspective) became, arguably, the world's number one emigrant nation, and the number one immigrant group in the US, Canada, Australia, and New Zealand (as well as some more lightly settled imperial holdings).
Meanwhile, contrary to well-worn ideas regarding Britain's small state apparatus and supposedly light taxes, in reality she used high taxation to build one of the biggest state infrastructures the world had ever seen, buoyed by prodigious military spending, and on top of all this the world's largest empire.
And while most 19 th century British emigrants went to the United States rather than to territories under direct British control, we argue that much of this flow to the US was still, for all intents and purposes, a story of British settlercolonialism-just not under the control of the British state.At the very least, it was to a significant extent 'settler-colonialism by British diaspora'.After all, this wave of British migration coincided with the USA's dramatic expansion westward, and the English-speaking settlers who carried this out were probably not, by and large, the descendants of Mayflower-era pilgrims, so to speak, but people of much more recent British extraction.It would have been numerically unlikely for this to be otherwise: prior to the 1820s or so, migration to the US/13 colonies (from Britain or anywhere else) had been little more than a trickle, and it was only in the decades following independence that immigration swelled the US population to potentially continent-dominating proportions.Descendants of the early pioneers, few in number as they were, could probably not have 'won the West' on their own.Indeed, far from putting a stop to British colonisation of the Americas, it was US independence that, quite paradoxically, launched this process of colonisation into its highest gear.This was no mere coincidence: the British government policy of arresting further westward expansion of the 13 Colonies had been one of the major grievances motivating the American Revolutionary War, and independence was instrumental in shaking off this constraint and decisively opening up the frontier. 2oreover, the combination of prodigious emigration, settler-colonialism, and imperial state expansion on this scale was peculiar to 19 th century Britain, and we suggest that this peculiarity helps explain Britain's divergence from the psi patterns that pushed other European countries toward revolution.Even France, with a reputation for 19 th century imperialism and colonialism to rival that of Britain, never saw emigration rates anywhere near those of her old enemy.
Crucially, these intertwined experiences of emigration, settler-colonialism, and imperial state expansion address both the popular (mass mobilization potential (MMP)) and elite (elite competition) prongs of Goldstone's psi measure, in that they provided outlets from Britain's structural-demographic pressure cooker for people from a wide array of social positions-from the unskilled labourers who filled the troop ships to the empire or the steerage compartments on transatlantic crossings to the Oxbridge graduates who found postings as administrators in far-flung colonies or as army officers.
Clearly, the idea that imperialism and settler-colonialism might provide such a social 'safety valve' is not an entirely new one.However, it figures surprisingly little in the sometimes rather abstruse literature in which rival theories of imperialism and colonialism do battle.Systematic attempts to link any such notions to the study of revolutions are similarly rare.Rarer still is any attention to the concept of an 'elite safety valve'.That is, the 'safety valve' concept commonly evokes notions of excess numbers of 'poor people' (however defined) in society.But we know from the work of Goldstone and others that revolutions are rarely, if ever, caused by an excess of poor people alone-and that as or more potentially destabilizing is an excess of elites, elite aspirants, and downwardly mobile people struggling to retain elite status.We suggest that 19 th century Britain had certain of these groups in spades, and that the rapidly expanding state, officer corps, and empire were instrumental in diverting their energies into channels that did not threaten societal stability.
Thus, while Goldstone stresses the role of an expanding and industrializing economy in absorbing 19 th century England's demographic shocks, we accept this analysis but argue alongside it for similar attention to the vector of emigration, settler-colonialism, and imperial state expansion into which at least some of the exhaust fumes of the population explosion were vented.Furthermore, it is important to note the crucial role of a highly interventionist state and 'big' government in the background to these dynamics-a far cry from the light-touch, laissez-faire qualities with which the 19 th century British state is often associated.
To make our case, this article takes advantage of secondary literature and raw data not available prior to the publication of Goldstone's book.Of crucial importance here is our unique dataset of fatality-inducing political violence events in Britain and Ireland from 1785 to 1900.This is the first research paper to utilise this dataset.We draw upon this in the following section, which seeks to establish what the real level of political instability was in 19 th century Britainthus cross-referencing Goldstone's account with more recent data-before moving on in the following section to a more detailed overview of the socioeconomic conditions underlying events at the political level.This is followed by our account of the emigration-settler-colonialism-imperial state expansion vector and the interventionist state policy behind it, which we argue was crucial to making 19 th century Britain relatively 'revolution-proof'-alongside the expanding economic opportunities rightly highlighted by Goldstone.Lastly come our brief concluding remarks, which lay out the implications, as we see them, of this article's findings for research on revolutions, political violence and instability, demographic-structural theory, state-building, migration, and imperialismcolonialism.
In making our case, we shift the main unit of analysis from England, upon which Goldstone focused, to that of Britain as a whole.While Goldstone's analysis of England is a crucial reference point in this article, we believe our choice is justified on a number of grounds.Goldstone was concerned with the relationship between relative political stability and the dramatic economic transformations most visible in certain parts of England-but for us, this is part of a wider British story, and we believe that this is borne out by the findings of our original data analysis, which show similar patterns of political violence (which we use as a proxy for political instability) across Britain as a whole.Furthermore, our argument focuses more than Goldstone's does on state policy in this period, and this entails a British, rather than English, focus-even if England formed the centre of gravity of the British state, just as much as it did the British economy.To speak of the British state in turn brings up the thorny subject of Irelandformally just as much a part of this state as England, Scotland or Wales, but at the same time clearly apart.Ireland does form an important part of the story of political stability and instability in the 19th century United Kingdom, but the findings of our data analysis underline the extent to which Ireland was an outlier from British norms.Ireland was part of the UK, but it was not Britain-a fact which would fuel much of the political violence seen in our data analysis.

What Happened?
Goldstone's account of Europe's 19th century revolutions involved a calculation of political stress indicator (psi) scores for England, France, and pre-unification Germany between 1820 and 1847, and their correlation with obvious sociopolitical crises such as those climaxing in 1830-32 and 1848.Goldstone is mainly concerned here with the socio-economic context of stability and instability, as opposed to the details of that instability itself.To extend his work and to get a measure of political stability that would span the period, we conducted a novel analysis of primary sources at Trinity College Dublin (TCD) in order to measure the frequency of fatality-causing political violence events between 1785 and 1900.However, before looking at the TCD data-focused as it is on fatalities-it is useful to look at an alternative attempt to systematically gather data on popular contention in Britain in this period.Using a much broader definition of 'contentious gatherings', Charles Tilly collected data on a wide spectrum of extraparliamentary political and social movement activity in Britain between 1758 and 1834-an imperfect match for us, both in terms of periodisation and unit of analysis, but entailing sufficient overlap with our parameters of interest to be of considerable use.The idea here is to use the TCD data, cross-referenced with Tilly's, to go beyond Goldstone's somewhat schematic picture of political instability in 19 th century England and France, and ask: how much political instability was really occurring in 19 th century Britain, and what did it look like?

Political Instability in England and France During the 19 th Century
As with his early modern cases, Goldstone made psi calculations for 19 th century England, France, and Germany on the basis of mass mobilization potential (MMP) multiplied by elite competition-using food prices and real wages as proxies for the former, and university and preparatory school recruitment as proxies for the latter (more details of which below).Unlike his early modern cases, he did not factor fiscal crisis into his 19 th century psi calculations, arguing that this was rendered irrelevant by its absence-with 19 th century rulers having learned to consolidate their fiscal structures so as to ward off 1789-France-type fiscal collapse at all costs.
As can be seen in Figure 1, the psi curves for England and France in this period (which he compares separately from Germany) look similar, except that the peaks are less pronounced in England and they taper off after a mild peak around 1841, whereas in France the curve keeps climbing towards the pivotal crisis point of 1848.England's curve does start to rise again around 1846, but this is from a very low base.France and England, 1820-1847. After Goldstone (1991: 312).

Figure 1. Pressures for crisis (psi) in
A comparison of the pressures for crisis in England and Germany yields a somewhat different picture, but nonetheless reinforces the apparent divergence between England and at least some parts of continental Europe in the lead-up to the revolutionary year of 1848.Thus, although Germany shows less dramatic peaks and troughs than England or France for much of the period in question, we see Germany's psi curve climbing sharply between 1845 and 1848-the latter as tumultuous a year in Germany as in France-as England's remains extremely low (albeit starting to rise).England and Germany, 1820-1847. After Goldstone (1991: 341).Goldstone (1991) explains the crises of these years as having much in common with the earlier cases of revolution and state breakdown featured more prominently in his seminal account.Disagreeing with the view of Tocqueville and others that these were not 'real' revolutions, Goldstone saw them as bona fide state breakdown episodes (or perhaps 'nearly state breakdown' episodes).However, he found the Marxist explanations of these events in terms of class struggle to be similarly lacking, arguing that members of the same classes could be found on both sides in these revolutions.For Goldstone, these 19 th century battle lines were drawn along "factional rather than class divisions" (Goldstone 1991: 287) and, rather than arising in response to 'the new', or in other words 19 th century capitalism, they represented 'the old'; that is, they were the last round of the species of demographic-growth-induced revolution associated with the early modern era.

Figure 2. Pressures for crisis (psi) in
Furthermore, they occurred where development lagged behind, rather than where it was precocious.Thus, while old-fashioned state, fiscal, and class structures were unable to deal with 19 th century demographic pressures, regions more developed in capitalist terms were better able to deal with these pressures thanks to the new resources and opportunities that 19 th century capitalism brought about.
The great difference between the 19 th century and the early modern era was that 19 th century demographic pressures were much greater.This trend, seen widely across Europe, was most extreme in the UK, where population grew by an unprecedented 92% in England and Wales (treated as one unit for many administrative purposes) between 1800 and 1850, and 72% in the UK as a whole (Goldstone 1991: 290).German population grew 61% between 1816 and 1864most intensely in the north-while French population increase, although not quite as fast, still registered an impressive 31% from 1800 to 1850.
Goldstone underlines that some regions could deal with population growth better than others.Thus, while France was seeing significantly lower population growth than either England or Germany, at the turn of the century she was still experiencing the kinds of pressures on grain that marked the most troubled years of the Ancien Regime.Agricultural output was just about keeping pace with population growth on average, but in bad years it fell behind, leading to dramatic spikes in grain prices, and "crisis peaks in 1830 and 1847" (Goldstone 1991: 291)-as illustrated in Figure 3:  Goldstone (1991: 303).Like Germany's overall psi curve, German grain prices did not display swings as dramatic as those of England or France-although 1847 saw a slightly higher peak than in England, while significantly lower than that of France.This can be seen in Figure 4: After Goldstone (1991: 291).
Prices of staple goods are more meaningful, of course, when viewed in relation to wages.The period in question saw considerable fluctuation in wages, as can be seen in Figure 5.It is worth quoting Goldstone at length here: In all countries, the years to 1827 were prosperous for workers; thereafter a severe crisis struck France, and milder drops in income occurred in 1830 in England and Germany.There was recovery in the mid-1830s, but in France and Germany wages then steadily drifted downward, falling precipitously in the crisis of 1847-1849.By contrast, though English wages show ups and downs after 1830, the trend in real wages is flat and sustained at a higher level than in the 1820s.The overall picture is one of improvement in England, admittedly with some ups and downs and some stagnation in the 1840s, but one of increasing difficulty in France and Germany, with particularly sharp periods of misery in 1828-1830 and 1847-1849 in France (Goldstone 1991: 293).After Goldstone (1991: 293).Note: Real wages are shown as an index, adjusted so that for all countries, real wages in 1820-1820 = 1.00.
The wage and price movements shown above formed the basis of Goldstone's measure of mass mobilization potential-one dimension of psi, which, as we know, pointed toward significant social strife in England and France around 1830-32, and in France and Germany around 1848.Most important for Goldstone, of course, was causal explanation of the forces driving these dynamics, and this pointed ultimately to demographics.The rapid population growth was palpable in 19 th century England, France, and Germany, with patterns familiar from earlier periods of such growth clearly visible; thus while real wages fell amid "saturated labor markets", rapid urban expansion saw Paris doubling in population from 1800 to 1850, as London nearly tripled.Regional cities experienced similar growth, with Lyon and Marseilles almost doubling, and Toulon tripling.Meanwhile, urban infrastructure and services (including policing) became severely overstretched (Goldstone 1991: 292).
However, as Goldstone (1991: 293-294) made clear, "Urban misery... is not enough to provoke crises, unless elites unleash the high mass mobilization potential of aggrieved workers through their own attacks on the state."Such elite unrest constitutes the second dimension of Goldstone's psi calculations for this period, which he frames in terms of 'elite competition'-that is, people competing to attain or retain the finite social resource that is elite status.Goldstone measures this via the proxy of enrollment rates in universities and elite schools, which, as scholars such as Stone (1974a andb) andCollins (1981) have long noted, display distinct boom-bust cycles over the medium-to-long term, closely tracking trends in social mobility, status anxiety, credentialisation, and credential inflation.Of course, rather than necessarily 'solving' any of these social tensions, booms in university enrollment often add fuel to the fire, by turning out cohorts of graduates who do not necessarily find appropriately 'elite' posts to match their qualifications and expectations.Goldstone points out that challenges to the state by elite aspirants "are particularly likely when the number of aspirants to elite positions is growing faster than the number of places" (Goldstone 1991: 294), and that this was certainly the case in the early nineteenth century: ... there were employment opportunities galore on the bottom rungs of educational ladders, where the vast expansion of popular journals and schools provided low-level points of entry to the professions.But there was no such expansion in the upper reaches of state office, university life, and law and medicine.Ever larger cohorts of youth left the schools and universities only to find themselves crowded into the outer reaches of the bureaucracy and the professions... (Ibid).
Goldstone notes that, while these dynamics could be observed in all three countries in question, they varied in degree, timing, and effect.France saw enrollment in lycées (elite preparatory schools) almost double between 1816 and 1830 alone, before climbing again by half up to 1848.University enrollment in Germany tripled between 1800 and 1830, while in England, Oxford and Cambridge (still the only English universities at the start of the 19 th century) ended a long period of virtual stagnation by seeing enrollment double between 1800 and 1820, sustaining that rate into the 1840s, and then increasing enrollment again in the 1850s.
In France and Germany, it soon became clear that neither the economy nor state institutions were expanding quickly enough to provide sufficient or satisfactory positions for the ranks of freshly minted graduates.But in England, according to Goldstone, development outstripped that seen in France or Germany: the foundations of the modern civil service were beginning to be laid, and after some difficult years, the rapidly expanding economy would eventually offer considerable employment opportunities-even providing alternative, university-eschewing career routes for ambitious but less scholastically inclined individuals (Goldstone 1991: 294).
This distinction between the 'aspirant elite absorption capacity' of England and the continental countries is key to Goldstone's explanation of their divergence in terms of stability and instability as the mid-19 th century approached.In his account, the frustrations of insecure and marginal members of the elite world figured prominently from the first stirrings of unrest in the crises of 1830-32 and 1848.However, he distinguishes between the elite competition dynamics surrounding these crises and those of the Great Revolution of 1789 in France-suggesting that the different nature of the elite dynamics involved help explain why the crises of 1830/1832 and 1848, though extremely serious, stopped short of the momentous proportions of 1789.
Thus, while the hopes, desires, and frustrations of aspirant elite members came into play in 1830/32 and 1848, 1789 had been marked by elite anxiety of a different, indeed almost opposite, nature: that of elite members in fear of the loss of status they already possessed.This gave to 1789 the potential for much greater violence, as a whole layer of such disgruntled and downwardly mobile elites were ready to fight to defend the status quo, or at least their idea of it.Indeed, a significant factor in the dynamics of 1789 was the presence of an element of the old nobility who wished to use the crisis to bring about a return to the receding status quo ante in which their privileged position had not been challenged by the noblesse de la robe and others (Goldstone 1991: 314).But the 1830 and 1848 equivalents of these elite elements were (somewhat ironically) more stable and assured of their social positions than had been their forebears in 1789, and thus were not willing to put up bloody resistance in defence of the status quo.This was also partly because these elites had learned to manage their birthrate in the 19 th century-thereby limiting competition anxiety-and partly, Goldstone seems to suggest, because many elite elements were not prepared to get shot defending men who had ruled so badly-thus rendering even militia units drawn from wellto-do segments of the population unreliable (Ibid.).
Thus, we can sum up Goldstone's view of the revolutionary crises of 1830 and 1848-at least in France-as something like serious, but milder aftershocks of the kind of chronic demographic strain that ultimately brought about the Great Revolution of 1789.Marx was wrong about the origins of these crises: they were not the birth of the new and it was neither lumpenproles nor textile factory workers who manned the barricades-rather the barricades were manned (and broken) by members of the traditional crafts who were facing the crisis of falling wages and severe unemployment thanks to the inability of the system to deal with the demographic challenge.Meanwhile, members of higher social strata were in their own state of revolutionary ferment, as the same demographic forces put them under pressure via elite overproduction and ensuing elite competition.But as serious as these crises were, they played out in rather different political circumstances than those of 1789: compared to the Great French Revolution, incumbent elites had less incentive to fight tooth and nail for the status quo, and state fiscal structures had been stiffened against collapse (notably thanks to the reforms of Napoleon and Villèle in France) (Goldstone 1991: 313).In the absence of state fiscal distress, the conditions for major crisis were present, but what was not present was the widespread sense that the whole monarchical structure was so corrupt and ineffectual as to require violent, root and branch overhaul.Instead, a radical shake-up of the system was deemed sufficient.Thus, in Goldstone's view, France in 1830 and 1848 effectively skipped the bloodletting of 1789-1794 and jumped straight to the equivalent of the Directory-a more or less conservative populist dictatorship (Goldstone 1991: 315).

England, 1640-1832: From Revolution to Reform
The English case bore for Goldstone significant structural similarities to France, but also crucial differences.Indeed, in something of a recurring pattern, we see many of the same phenomena as in France, but in different order, and with different magnitude and significance.We have considered briefly Goldstone's observations as to how England came to diverge from France and Germany after 1830-1832.We will now take a step backward, and look in somewhat more depth at what had been happening in England and Britain prior to this divergence.
In a sense, England had had her 1789 more than a century before France, and since the Restoration had enjoyed greatly eased social pressures in the late 17 th and earlier 18 th centuries: population and prices were stable, real wages were increasing, labour markets were tight, and social mobility had declined.Indeed, elites had lowered their birth rates-sometimes being left without male heirs and having to merge two family fortunes together, to absorb promising professional commoners, and so forth.Reduced elite competition was reflected in reduced admissions to Oxford and Cambridge, the total of which had actually fallen from 378 to 321 per year between the 1740s and the 1760s (Goldstone 1991: 328).To put this in perspective, the average annual freshman admissions to Oxford alone had been 400 in the decade 1620-1629 and 530 in the subsequent one-the period of intense elite competition presaging the revolutionary crisis of the 1640s-if Stone's (1974a: 91) calculations are correct.
However, the remarkable stability of roughly 1650 to 1750 was not to last.Around the middle of the 18th century, England entered into the breakneck population expansion that would become known as the demographic transition, and that would ultimately change the country, and the world, forever.Goldstone (1991: 323) points out that, after growing by only 10% over the previous century, the population of England and Wales nearly tripled between 1750 and 1850, jumping from 5.7 million to 16.5 million.Though this pace quickened after 1800, the 1750-1800 growth was substantial, with Anderson (1996: 211), based on slightly different calculations, suggesting that the population of England alone grew from 5.8 million to 8.7 million over those fifty years.
By the late 1760s, serious social strife had returned to England, recurrent waves of which would roil the political system until they reached their climax in the events of 1830-32.
These waves of unrest would take various forms and various banners in an increasingly complex political conjuncture, and would have both popular and elite dimensions.Underlying them, however, were all too familiar structural and material strains induced by rapid demographic shifts.By the 1760s, growth in agricultural output had fallen behind the rate of population increase; accordingly, wheat prices grew by 60% between the 1740s and the 1780s, and by almost 40% again between the 1780s and 1825-31, in spite of post-war deflation (Goldstone 1991: 324).Real wages fell and unemployment rose-especially in the south, where relative lack of industry compared to the midlands and north meant an over-supply of agricultural labourers.With smallholdings still abundant (one third of privately owned plots were less than four acres) and yields per acre stagnating, draconian changes in land use patterns were necessary to increase food production; but even so, food imports from Ireland (formally part of the UK since 1801, but effectively still a British colony and captive 'food ranch') became increasingly important to maintaining an adequate food supply.
These developments had brought about considerable misery at a popular level by the late 18 th century, particularly in the south.They also shaped sharply increasing tension and competition at the level of the elites.The dramatically increased numbers of university graduates being turned out by Oxford and Cambridge faced uncertain prospects in the late 18 th and early 19 th centuries: the civil service would not begin its major expansion until the 1830s, and the Anglican church-traditionally a mainstay for Oxbridge graduates-was hamstrung from expanding to meet the latest population increases due to internal bureaucratic and institutional inertia.Meanwhile the non-Anglican population was starting to figure much more prominently in UK public life: the Catholics of Ireland and of the expanding Irish diaspora in Britain, and the rapidly growing dissenter congregations of Britain, were producing their own elites amid the new economic circumstances of the era, and these were demanding inclusion in a British political system that had been designed to keep them out.
In the complex political battlefield of late 18 th and early 19 th century Britain, these movements produced considerable backlash, with anti-Catholic campaigns leading to major bouts of unrest-particularly the bloody riots of 1780.However, to some extent these calls for the opening up of the system to non-Anglicans dovetailed with wider popular demands for reform of a system seen as corrupt and landlord-dominated.Not unreasonably, the Corn Laws fuelled a widespread sense of landlord conspiracy to keep grain prices high.This naturally fed into one of the other great political hot button issues of the day, namely the continued existence of 'rotten' or 'pocket' boroughs-archaic micro-constituencies with tiny handfuls of electors, easily bought off and controlled by local members of the landlord class in order to secure safe seats in parliament.The 'rotten' boroughs were sited in depopulated ghost towns that had once been major population centres-and the rapid emergence of new, bona fide population centres in industrial cities such as Manchester and Birmingham, without their own parliamentary representation under the old electoral maps, added to the pressure for reform.
Thus, a popular-marginal elite coalition for reform of the closed, landlorddominated political system gained ground.Major bouts of political violence, such as the massacre of repeal supporters by yeomen at 'Peterloo' in 1819, and the 'Swing' riots of agricultural saboteurs in 1830, gave legislators an added sense of urgency.An increasingly dangerous situation was eventually defused in 1832 with the passing of the Repeal Act-a long way from a full democratic charter of equal rights for all, but a decisive step in the direction of a somewhat more democratic and less corrupt electoral system, and a modicum of what would now be considered citizenship rights.After a number of blocked attempts to move such a bill since 1830, William IV had stepped in to call a new parliament in 1832, and it was this body, with a strong reforming majority, that passed the Reform Act.Goldstone (1991: 332) speculates as to what might have happened had William IV acted differently: If William IV, like Charles X, had reacted to the crisis by dissolving Parliament and failing to call for new elections, and had he backed an unpopular conservative ministry and imposed censorship, it seems likely that the Whig leaders and middle-class reform opponents might have taken their protests to the streets.In this event, would the army have fired on mass demonstrations led by respectable citizens against an unpopular government?Would William have been forced to abdicate?Might there have been a revolution in 1832-not like that of 1789 in France, but like that of 1830-in England?
Clearly, Goldstone credits William IV and his government with acting prudently in successfully de-escalating the 1830-1832 crisis.But in explaining the further divergence of England from continental patterns from that point in time onwards, he also highlights the importance of England's much more fortuitous material circumstances-conditions that, given relatively sane governance, made it possible for England to deal with challenging demographic forces that would continue to founder similarly afflicted continental neighbours for decades.For, while Britain's population growth far outpaced that of France and Germany in the early 19 th century, so did her industrial progress-which put her 30 years ahead of France if measured in terms of aggregate steam horsepower, and 40 years ahead of Germany (Goldstone 1991: 325).Indeed, in 1840, Great Britain had 72% of Europe's total steam horsepower: 620,000 as opposed to 90,000 in France.This disparity is all the more striking when relative population is taken into consideration-with that of England, Scotland, and Wales together still making up less than half the French population in 1830 (Ibid.).Thus, Britain had 33.3 horsepower per 1000 inhabitants, compared to France's 2.6.Likewise, trade and industry were already a bigger part of the British economy than the French as early as 1789.By 1840 only 22% of England's labour force was engaged in agriculture, which accounted for one fifth of UK national product, whereas in both France and Prussia the labour force engaged in agriculture was still over 60% in 1850, and in 1830 French agriculture was still half of all national output (Ibid.).With respect to multiple industrial output/consumption milestones (tonnes of pig iron produced, cotton spun, etc.), England/Britain put herself 30-40 years ahead of France and Germany.
These statistics matter because without the industrial jobs they made possible, the society could not have coped with the new demographic reality, which would otherwise have brought untold misery.Thus, Goldstone again emphasizes his inversion of Marxian thinking on this topic: the problems of the era were not caused by proletarianisation or by the Industrial Revolution; rather, given the population explosion in progress, the huge glut of new industrial jobs was the society's saving grace.Indeed, it was the regions lagging behind in industrialization-such as the previously more affluent south of England-that suffered most in this era, while workers in the new industrial heartlands to the north benefited from rising wages and relatively stable employment-rather than having to desperately 'auction themselves off in the satanic mills,' as it were.
Thus, Goldstone's explanation of England's exit from the cycle of demographically-induced revolutionary crisis after 1832 hinges on a combination of prudent parliamentary reform at a political level, and the continuing rollout of industrial development, and corresponding employment opportunities, far above and beyond that of her continental neighbours.
We will briefly come back to this point, and enquire as to whether this does indeed suffice to explain Britain's divergence from France and Germany after 1832.But first we will address a distinct, but important point: the question of how much and what kind of political instability really was occurring in Britain at this time.

Measuring Political Instability and Violence in Nineteenth-Century Britain
Goldstone's psi calculations yield a vivid picture of the pressures building towards political instability and violence, but they cannot measure that political instability or violence directly.Thus, we look here at data collected with precisely that aim in mind.These show us that, if anything, Goldstone may have understated somewhat the level of unrest afflicting English society in this period.
Tilly's data overlap imperfectly with the periodisation of our study, but they do cover the crucial years from 1820 to the early 1830s.They do not cover every year within this period, and they are collected from a limited range of contemporary periodicals-covering only the south-east of England up to 1828, and the south-east of England plus the rest of Britain from 1828 onwards.
The headline features of Tilly's data can be summed up as follows.What he classifies as 'contentious gatherings' (CGs)-everything from public meetings to pitched street battles-saw upsurges around 1819/1820 and subsequently much more dramatically from 1830 to 1832-peaking in 1831 at around 18,889 CGs (see Figure 6 and Table 1) or around 250 CGs per million of population.As indicated above, CGs represent a catch-call category in Tilly's data, taking in many different kinds of events, and thus these patterns could be interpreted in many different ways.Indeed, much of Tilly's argument is devoted to tracking the changing nature of 'popular contention' in Britain over time, qualitatively as well as quantitatively, which he argues shows a distinct evolution from small-scale, localised, raucous, and often what we might call 'popular rough justice'-type CGs in the late 18 th century to the emergence of the much more formalised, standardised, sober, and large-scale modern social movement form in the early 19 th century.Thus, since a turn-of-the-century CG is not necessarily very comparable to an 1830s CG, it is useful to look also at some more directly comparable measures of contention in Tilly's data: those relating to numbers of deaths, woundings, and arrests associated with CGs year by year.We can see from Table 1 that the highest casualty figure for the 19th century years in Tilly's sample comes from 1831, with 52 deaths recorded (4 in "south-eastern England" and 48 in the "rest of Britain").The second highest figure among the 19 th century years is for 1807, which records 32 deaths for the south-east of England alone-data for the rest of Britain being absent from Tilly's sample for this year-although on closer inspection, most of this year's deaths have little to do with the kind of political violence and instability we are concerned with here. 3The third highest 19 th century death toll (28 fatalities) in Tilly's sample is for 1832-a year high in contention for obvious reasons, although not accounting for as many deaths, if these figures are anything to go by, as the previous year.It is also worth noting that while Tilly's data record only 5 deaths for 1819, again this is for south-eastern England only, and if the rest of Britain were included in the sample for this year, the figure would be considerably higher-this being the year of the "Peterloo Massacre".Though not included in Tilly's sample, he suggests that perhaps 15 people died at "Peterloo" (a grimly comic reference to 1815's Battle of Waterloo), when armed yeomen set upon a crowd gathered at an open-air meeting at St. Peter's Field in Manchester, calling for parliamentary reform and repeal of the Corn Laws-although the exact death toll is uncertain (Tilly 2005: 80).Unsurprisingly, arrests and woundings are also heavily concentrated in the years around the Reform Crisis-with both types of incident peaking in 1830, followed by 1831 and 1832 in second and third place in either case.Nonetheless, the most violent of these 19 th century years do not approach the levels of violence seen in the 18 th century part of the sample.Thus, the 109 deaths of 1780 and 75 deaths of 1768 surpass the violence of the late 1820s and early 1830s, even though the earlier figures are for SE England only, and spring from a lower total population. 4This underscores Tilly's point that popular contention in Britain became less violent over the course of his sample, at the same time as it became more frequent-with the watershed between these two eras of popular contention ("repertoires of contention" in Tilly's verbiage) seeming to come sometime toward the end of the Napoleonic wars.More specifically, "... violent sorts of events declined from three quarters of the total in the 1750s to around one-tenth in the 1830s.Meetings of one kind or another rose from 15 or 20% to over 80% of all CGs... [and] dominated contention from 1807 onward (Tilly 2005: 96)."This is illustrated graphically in Figure 7: Tilly points out that the key to this downward trend in violence may lie not just on the side of those doing the "popular contention."As people became less likely to adopt violent tactics, and as their activities became more decorous (public meetings and respectable marches, etc.), things changed on the side of the state also: as policing became more professionalised and institutionalised, it became less likely that troops would be sent in against crowds.This still happened in the 19 th century, as seen at Peterloo-but not on the scale of the 18 th century, as exemplified by the notorious bloodletting surrounding the Gordon Riots of 1780 (Tilly 2005: 94-95).
Tilly's data also reveal the sheer diversity of the issues motivating popular contention in this era.In coding his sample of approximately 8,088 CGs, Tilly identified about 3,000 different issues; however, 77% of these can be accommodated in seven broad (and, alas, rather vague) categories: attack on a person or object; religious issues of any kind; elections; parliamentary reform; misery; government; and labor.The incidence of the first four of these categories is graphed in Figure 8-revealing no clear long-term trends, but considerable short-term fluctuations in the salience of particular issues.Tilly describes the curve for 'parliamentary reform': To be sure, our scattered years miss the grand reform debates of 1782-1785, 1790-1793, 1809-1810, 1816-1817, and 1822.Yet they capture the steps up from 1780 to 1811 to the incomparable heights of 1831, when at least 953 of the 1,645 CGs in our collection-58%-concerned reform, and another 304 events (18%) concerned either elections or the national government.Like reform, many other issues ran in surges, building in a few months from low levels of action to high intensity.That tendency became more pronounced with the C19 transformation of repertoires, as associations formed, existing groups mobilized, public meetings proliferated, and leaders competed for the attention of Parliament as well as for the allegiance of other activists (Tilly 2005: 102).Tilly also addresses wider issues of data reliability, and suggests that any holes in the data are not fatal: ... the gaps in the chronology before 1828 could hide some greater year-to-year swings in contention as well as an earlier start to the post-1811 rise.Although among years of major subsistence crisis the sample includes 1758, 1768, 1795, 1801, and 1811, it entirely misses the crises of 1772-1773 and 1816-1818.During the period of the French Revolution the sample years skip the Church and King riots (1791) and a wave of attacks on recruiters and press gangs (1794).But the selected years do include major struggles over John Wilkes (1768-1769) and Lord George Gordon (1780), crucial years of the French Revolution and the wars with France (1789France ( , 1795France ( , 1801France ( , 1807France ( , and 1811)), the time of Peterloo (1819), and the great outcry over Queen Caroline (1820), as well as the continuous record of contention from 1828 through 1834.All in all, it seems unlikely that missing years are distorting general trends in the data (Tilly 2005: 89-90).
After addressing the need to adjust for population change, he also discusses how the data might be affected by changing levels of reporting: Increased reporting accounts for some (but surely not all) of the apparent rise of public contention.It is possible, if rather unlikely, that trends outside the SE ran in a quite different direction before 1828.These curves, furthermore, do not establish that C19 Britons made claims more often than their C18 counterparts; small-scale and indirect claim-making probably declined significantly after 1800, and could have dropped enough to compensate for the rise in larger-scale, direct claim-making.The evidence in the table and graphs simply indicates that gathering in substantial numbers and making public claims became a much more frequent way of contending in Great Britain after the Napoleonic Wars than it had been earlier.But that was an important change (Tilly 2005: 90).Indeed, because of the strict standards of Tilly's data collection, Tilly's casualty figures tend to run lower than in "conventional" histories, using as he is only data points his team could verify according to their sources and methodology.Thus, for example, their figures for the 1780 Gordon Riots are 106 casualties as opposed to a widely cited 285; 16 wounded vs. hundreds; and 161 arrests vs. 450 (Tilly 2005: 91).Likewise, in the rural rebellions of 1830, 1,327 arrests were verified by Tilly's team, as opposed to the 1,976 reported elsewhere.As for woundings, 1768 is the worst year in absolute terms-while 1780 (the year of the Gordon Riots) is surely under-reported (Ibid.).
To what extent are Tilly's data from southeastern England indicative of trends beyond that immediate region?Does the absence of certain years in his sample obscure important trends?Does the rise in popular contention visible by 1819 start much earlier, during the post-1811 years missing from Tilly's sample?And, perhaps most fundamental of all, how well do Tilly's data stand up in the light of further research?Has he significantly understated or overstated popular contention or political violence during the period in question?

Methodology for Constructing TCD Political Violence Dataset
In order to answer these questions, and to provide us with greater insight into the dynamics of political instability in Britain during the period in question, we followed Turchin's methodology (Turchin 2012) to create a dataset that would robustly capture the dynamics of political instability in these years (covering the 1780s to 1900).The idea is that we only include events in our dataset that are both clearly political in nature and involve fatalities, as these are the events that are least likely to be omitted from patchy historical records.We analysed contemporaneous newspaper accounts to identify reports of political violence leading to fatalities and count their frequency.Because we are primarily concerned with the dynamics of instability, the consistency of our source is more important than its objectivity or veracity; we are not trying to measure how many people got killed in political conflicts, we just try to measure how often the newspapers report people getting killed in political conflicts.
For the purpose of this structural demographic analysis, we are only interested in reports of internal political violence, not external wars or reports of foreign massacres.Throughout the 19 th century, Ireland was part of the British state5 and, thus, political violence in Ireland is included in the domestic tally.However, events in Ireland and the rest of Britain were counted separately, as basic historical knowledge of their different trajectories in the 19 th century would suggest that they might diverge strongly during the period.
In this case, we are fortunate to have a single source that covers almost the entire period in a consistent manner: The Times newspaper began publishing in 1785 (albeit under the title The Daily Universal Register until 1788), and continued to publish daily throughout the 19 th century, without any noticeable deviation from their unwavering support for the empire and the order that supported it.Their full historical archives are available electronically on the web at: http://www.thetimes.co.uk/tto/archive/.
Unfortunately, the Times archives were available exclusively as scanned images, of a quality that made Optical Character Recognition (OCR) and automated natural language processing ineffective.Therefore, in order to identify such reports and filter out false-positives, we were limited to the search engine provided by the Times.We invested a significant amount of time in crafting and verifying search queries that would be broad enough to capture all of the reports that we were interested in, without producing an impossibly large number of scans to read.We created a gold-standard dataset for 1831, combining both human-expert and automated methodologies (Feeney 2014).This gave us a benchmark with which to tune search phrases that produced both high coverage and acceptable levels of false positives.It produced 49,935 results for the period in question.However, naturally, these results included large numbers of false positives-reports of non-political violence, political violence without fatalities, or just plain irrelevant events.
Search results are made available by the Times as scans of fragments of tightly-printed, age-worn newspaper pages and are often difficult and timeconsuming to read.They are also sometimes difficult to interpret, requiring significant domain expertise to understand the political significance of events.To limit the requirement for this scarce expertise we adopted a sampling plan, where we broke the data up into several distinct samples of approximately 1% of the data, with each sample containing a different selection of newspaper reports from across the time-period, and then used our gold standard and an analysis of their self-similarity to evaluate their robustness.We then averaged the results of each year by decade, in order to smooth out sampling anomalies.This reduces the granularity of the dataset-it only represents changes between decades, not individual years-but increases its robustness considerably.As we are interested in identifying long-term structural demographic processes rather than finegrained year-on-year changes, this tradeoff is acceptable.
We produced three samples, using different algorithms, each consisting of 739 scanned images and analysed them to identify reports of political violence events with fatalities, totaling 2217 scans.145 (6.5%) turned out to be about domestic political violence events with fatalities, the rest were false positives.The results were consistent between the samples.Extrapolating from the samples to the full dataset gives us the results show in table 3 and figure 9 below.Table 4 presents the results of our gold-standard analysis on 1831, which included several different methodologies-including the traditional approach whereby a historian reads and analyses every paper published in the year.The Times archives for 1831 were found to contain, in total, 97 reports of political violence events, which referred to 44 distinct historical events that took place in 1831 (2.18 reports per event) involving a total of 153 fatalities (3.48 fatalities per event) and one that referred to an event from 1819 (a retrospective on the Peterloo massacre of 1819 with detailed biographies of all 15 victims).34 (77%) of the events were situated in Ireland, with 112 fatalities (3.3 per event), covered in 66 reports (1.9 per event).10 (33%) of the events were located in Britain, involving 43 fatalities (4.3 per event), covered in 31 reports (3.1 per event).
It should be emphasized that 1831 is not meant to represent a typical year.It is used as a gold standard, both to assess the performance of our sampling and searching and the general accuracy of the data-source.It was a year of extraordinary turbulence in Britain, and included the Merthyr Rising of May and June 1831, which modern historians consider to be the political event that caused the greatest number of fatalities (24-26) in the history of 19 th century Britainwhat Davies (1984) referred to as "the most ferocious and bloody event in the history of industrialised Britain".Eight contemporary articles described this event.Unsurprisingly, they diverged on the number of casualties (varying between 8 and 29), but all agreed both as to the political nature of the event and that there were multiple fatalities.A second historically well-known political event with multiple fatalities that took place in 1831 was the Reform Riot in Queen Square in Bristol on 29 th October.This was covered in 6 reports in the Times, which again varied in terms of the details (estimating between 4 and 12 casualties with one wild outlier claiming between 370 and 380 victims) but not in terms of the political nature of the event or the presence of fatalities.
The frequency of events in the gold standard matches the predictions of our sample closely.If our samples are representative, we would anticipate an average of 43.1 events in the 1830s: 33.6 in Ireland and 11.5 in Britain.Our gold standard analysis of 1831 found 44 events: 10 in Britain and 34 in Ireland.They are not entirely independent predictions-10% of the sample for the 1830s was drawn from 1831 -but it gives us some confidence that the results that we draw from our samples will be at least ballpark accurate in scale as well as dynamics.

Interpreting the Results
The first and most basic conclusion that we can draw from our analysis is that, throughout the period between 1785 and 1900, Britain had generally low levels of political instability.The contrast with Ireland is particularly stark in demonstrating this.Even at the peak of political violence in Britain in the 1830s, the frequency of events causing fatalities was less than a third of Ireland's.For most of the period under examination, political violence events leading to fatalities were exceedingly rarely reported, and after the 1830s, they virtually disappear from the record.That is not to say that they did not happen, but that they were so infrequent as to not be picked up by our sampling, either because they were occurring at a frequency of less than 1 per year or because they attracted little press coverage (or both).
Qualitative historical analysis of the Times articles from throughout this period supports the conclusion that the absence of such events from the record is a real phenomenon, and not due to sampling bias.Over the course of the 19 th century, the amount of reporting per event increases dramatically-each time such an event occurs, it appears in parliamentary debates, court-case dispatches, and opinion pieces.Major events with multiple fatalities, such as those that took place at Peterloo in 1819, Merthyr Tydfil in 1831, and Newport in 1839 (Davies 1984) continue to be referred to in newspaper articles for decades.After the 1830s, where such events occur they always have small numbers of fatalities, and events causing more than 10 deaths disappear completely from the record.Furthermore, it was not just that political violence events causing fatalities were rare throughout this period, but that significant political violence events and murderous events in general were also relatively rare, and tended to attract increasingly large amounts of coverage.
Thus, while Britain entered the period after 1785 with a very low level of political instability, this level gradually built up during the 1810s and 1820s to reach a peak in the 1830s, before collapsing and remaining at a very low level throughout the rest of the 19 th century.These results support both of Tilly's major conclusions: that political instability peaked in the 1830s before declining rapidly, and that political protest in Britain was in the process of becoming less violently contentious in the long run-a process well underway even by the 1830s.In our research, Luddites and Chartists show up repeatedly in newspaper reports, but they are very rarely associated with events causing fatalities.
A second major finding is that, from a structural demographic point of view, and in spite of the formal unification of Ireland with Britain in 1801 as a single state, the two islands must be considered as largely separate societies with very different dynamics.Ireland was consistently far less stable and more violent than Britain.In 1831, The Times reported on three distinct (and now quite obscure) political violence events in Ireland, at Castlepollard, Newtownbarry, and Carrickshock, each causing more than 10 casualties; as many as were reported for Britain for the entire 19 th century.Although they declined in both frequency and scale after the mid-1840s, until the 1890s, political violence events causing fatalities remained relatively frequent in Ireland.
In the early 19 th century a demographic crisis was building in Ireland, which peaked in the 1840s with a major famine between 1845-48.As well as reducing the population by 25%, the famine had the side effect of opening a massive emigration route to the United States, which provided an escape valve for Irish demographic pressures that persisted until the 1980s. 6The easing of demographic pressures after the 1840s saw a significant decline in Irish political instability in the second half of the century, and although intense structural pressures remained-a bitter and violent "Land War" broke out in the early 1880s-it was considerably less violent and long-lasting than the "Tithe War" of the 1830s, and reports of contentious Irish political events involving neither violence nor fatalities also become increasingly common in the records.
Finally, it is important to emphasize the limitations of these results based on sampling the data.We can confidently say that Britain experienced a peak of instability in the 1830s, and after that remained very stable for the remainder of the 19 th century.Because such events became rare in the records after the 1830s, we cannot interpret small variations in the numbers as being meaningful-for example, the small increase in Britain from 0 to 1.3 during the 1860s could easily represent a sampling artifact.Similarly, a score of 0 does not mean that there were no recorded political violence events causing fatalities in a given period, but merely that any such events were so infrequent that no references to them occurred in any of the samples of Times articles examined for that period.

Emigration, Settler-Colonialism, Imperial State Expansion, and Stability in Britain after 1832
Now that we have a clearer picture of the level of political violence and instability present in 19 th century Britain-as well as its changing character over time-we can turn back to Goldstone's account of the structural forces driving the movements of the psi curve in this period.We can see that the data broadly accord with Goldstone's account-in particular, the uptick in political violence and instability in the years around 1832, followed by a tapering off in psi and a return to relative stability-while France, as we may remember, careens toward further instability in the 1840s and beyond.However, as indicated above, we argue that Goldstone's account is necessary but not sufficient to explain England's and Britain's stabilization after 1832.That explanation correctly identifies England's economic dynamism as well as, secondarily, the expansion of the civil service from the 1830s onwards, as crucial to generating the blue collar and white collar jobs necessary to keep any popular or elite discontent at manageable levels-while parliamentary reform worked as a political corollary of this process.However, there is reason to believe that these measures alone could not have sufficed to stave off serious socio-political dislocation.
The facts are stark.Again, it is worth remembering the degree to which population growth in England outstripped that of other European countries at this time-a growth of 92% between 1800 and 1840 (compared to 31% in France between 1800 and 1850), which pushed England's population by 1850 to almost three times what it had been just a century before (Goldstone 1991: 290, 323).England and Britain survived this explosive population growth-but they could not have done this by domestic economic dynamism and growth of the civil service alone.
How can we say this for sure?Of course we cannot, technically speaking, prove this, in so far as to do so would require entering a complex web of counterfactual questions which cannot, by their very nature, ever be definitively resolved.However, we suggest that there is considerable evidence that Britain relied also on other means, alongside the domestic economic dynamism and civil service expansion highlighted by Goldstone, to dissipate her growing demographic pressures.
In fact, Britain's stability at this time rested on an extremely precarious equilibrium; keeping a country with such explosive demographic growth from crashing was a delicate balancing act.19th century Britain's vital statistics may look impressive when it comes to horsepower, pig iron production, or overall economic growth-but it is worth bearing in mind the point made by Goldstone himself when he argued that the great debate over the standard of living in 19 th century England is wrong-headed.That is, Goldstone pointed out that this debate tends to hinge on misleading aggregate statistics, which average out the differences between different regions of England, and indeed Britain.This obscures the important and basic fact that at the same time as certain regions were booming, others were in profound distress. 7This explains why in 1811 and 1812, as formerly sleepy villages elsewhere in the North and Midlands were metamorphosing into industrial powerhouses, 12,000 troops were occupying parts of Yorkshire and Leceistershire in order to quell Luddite unrest (Tilly 2005: 136); and why in 1830, as the opening of the world's first major inter-city railway between Manchester and Liverpool made history, the Captain Swing disturbances of England's distressed agrarian south-east led to as many as 1,327 arrests (Tilly 2005: 91).This also points to the complex realities that lie behind aggregate statistics on real wages and other measures of economic progress.For while real wages were growing considerably overall, this masked drastically different growth rates across different sectors of the labour market, and particularly tepid growth for agricultural workers-as demonstrated in Figure 10.We argue that economic dynamism was not a silver bullet that could solve these problems in a straightforward way-and indeed, paradoxically, in certain respects economic dynamism even exacerbated some of them.For instance, increasing efficiency and rising productivity in agriculture meant that there was relatively less need for workers; that is, while the agricultural workforce was increasing in absolute terms, this increase lagged significantly behind the rate of population increase.This resulted in relative oversupply of agricultural workers, especially in the south of England, and considerably lower wage growth in that sector.It also led to the flow of workers from agriculture towards other sectors, thereby creating further pressures and imbalances within the system (Tilly 2005: 120).
But 19 th century England and Britain did survive these demographic pressures without a revolutionary crisis, and without a major bout of political instability after that of 1832; does this not mean that the expanding economy did contain sufficient slack to absorb these pressures?We suggest that the answer to this question is no, because if it had not been for a series of other major developments in 19 th century Britain that were quite distinct from domestic economic dynamism-albeit related in intricate ways-it is exceedingly difficult to imagine how British political stability could have reached the end of the 19 th century unscathed.This is where the nexus of emigration, settler-colonialism, and imperial state expansion (ESCISE) comes in.These phenomena are relatively well known in and of themselves, to greater or lesser degrees.But we believe that their deeply interconnected nature in 19 th century Britain, and their key roles in maintaining political stability, have been significantly under-appreciated-as has the role of the highly interventionist (rather than laissez-faire) state lying behind these processes.
Of course, this is not to say that these processes were entirely state-driven or top-down; in some, there was a complex relationship between state action and relatively 'spontaneous' or 'bottom-up' non-state action.This is well illustrated by the case of emigration.Emigration from Britain had been negligible during the Napoleonic Wars and, indeed, British emigration to the Americasoverwhelmingly the main destination for British emigrants-had averaged only 3,500 per year between 1600 and 1776 (Hatton and Williamson 2005: 31).This trickle became a flood in the years and decades following Waterloo. 8In fact, European immigration to the US increased more than 21-fold between the 1820s and the 1850s-from a decadal average of 12,847 in the 1820s to one of 275,458 in the 1850s.Meanwhile, immigration to the Australian colony of New South Wales increased 10-fold during same period, and immigration through the Canadian ports of Quebec and Montreal increased by an even higher multiple.The UK was the main source of both the North American and Australian mass migration, with the numbers departing from there rising from 12,510 in 1816 (the year after Britain's defeat of Napoleon at Waterloo) to 176,554 in 1856representing a more than 14-fold increase over four decades (Hatton and Williamson 2005: 31-32).UK and British emigration far outstripped that from elsewhere in Europe for much of the 19 th century, even after emigration from other parts of Europe had picked up after mid-century-as can be seen from Table 5: 8 This is not to say that British emigration to the Americas remained flat until the late 1700s or early 1800s.Bailyn (1986: 9) points to a wave of migration "remarkable by the standards of the time" in the 15 years between the end of the Seven Years War and the American War of Independence, during which perhaps 125,000 people emigrated from Britain and Ireland to America.However, the dramatic upsurge in emigration from the 1820s onwards still represented a paradigm shift against these numbers.
Country 1851-1860 1861-1870 1871-1880 1881-1890 1891-1900 1901-1910 European Thus, for much of the 19 th century, the UK was at the top of the European emigration league table by migration rate per 1,000 of population-albeit being surpassed by Norway in the 1860s and 1880s, and by Portugal and Italy in the 1890s.On the aggregate-and considering the relative population difference between these countries-the UK was by far the leader in absolute terms.Indeed, between 1846 and 1850, the UK accounted for a full 78% of European emigration overseas (Hatton and Williamson 2005: 406).But this does not quite capture the significance of these figures.When juxtaposed with roughly comparable Western European countries such as Germany, the Netherlands, and France, we see UK emigration rates not just exceeding them, but exceeding them by large multiples.This is starkest with regard to France: as we see in Table 5, the French emigration rate hovered between 1.1 per thousand and 3.1 per thousand during this period-the lowest of any of the countries sampled.This means that for much of the period in question, UK emigration surpassed that of France by a factor of more than 50 to 1(!).In attempting to divine why patterns of political instability and the capacity to manage demographic pressures diverged in France and Britain after 1832, this curiously overlooked fact is surely worthy of some consideration.
What can explain this dramatic upsurge in emigration in the decades following the Napoleonic Wars?Specialists working on this topic have highlighted numerous factors, including rapid population increase in Europe, sharply declining transport costs (after they had skyrocketed during the prolonged period of warfare just ended), and the promise of significantly higher wages in the New World (Hatton and Williamson 2005: 34).Without a doubt, these factors, and others, sparked the desire of thousands of Europeans to vote with their feet, as it were, and strike out for a better life overseas, in a way that they could not have afforded to do before.
However, there was another crucial factor in explaining the opening of the emigration floodgates at this time: state policy.Several Western European states lifted restrictions on emigration between the 1820s and the 1840s-including the UK, which repealed separate emigration-constraining laws in 1825 and 1827 (Hatton and Williamson 2005: 40-42).9However, in the UK's case, there was a strong positive side to pro-emigration policy also.As well as lifting restrictions on emigration, the UK state was prepared to commit significant resources to organising and subsidising assisted migration overseas-most notably to Australia and New Zealand.About half of the 19th century mass migration to Australia and NZ-almost 3 quarters of a million people-was achieved by such means.From 1832-1851, the assisted share of migration to New South Wales was even higher, at 75% (Hatton and Williamson 2005: 42).
These were not insignificant sums: transport to Australia still cost about six times as much as the fare to America in the mid-19th century, and after assistance, the cost of a trip for a childless couple would still have been close to 20 pounds, or about two thirds of the average male farm laborer's annual earnings as of 1849-and even more for lower-paid Irish labourers.(The Irish accounted for about half of the assisted Australian immigrants in the 1839-1851 period).Such assistance probably cut the total cost of the trip approximately in half in the late 1830s/early 1840s (Hatton and Williamson 2005: 43).
However, there were major differences in the contending visions of assisted migration in this era, and major disputes over who would pay-in this case the UK government, or the Australian colonial governments.Consideration of the issues at stake reveals much about the role and significance of migration at this time.
Australia's colonial government sanctioned free immigration to Australia at the end of the 1820s, having relied on British convict labor until labor supply constraints made it difficult to exploit the European boom for wool exports.
Nonetheless, the question of who would be assisted to migrate there-the poor or the not so poor-was hotly debated at this time.Australians wanted to cherry-pick/quality-control their immigrants, while the prevailing idea in Britain was to use Australia (and New Zealand) as a dumping ground for pauperssimilar schemes in Canada and the Cape Colony having already failed.This disagreement was ultimately settled via the related debate as to who would foot the bill for the assisted passage scheme: the initial idea in Britain had been to finance it from a similar pool to that of the Poor Laws, but the idea changed to that of creaming off profits from sales of government lands in Australia; since this was seen as 'Australian money', it was leveraged by the Australian lobby in favour of their design for high quality control, and in 1831 they got their way.The significance of this is that, even if ultimately the Australian colonial government footed the bill for the assisted migration, there was considerable enthusiasm within British ruling circles to bankroll the use of emigration and settlercolonialism as a dumping ground for excess population (Hatton and Williamson 2005: 42-45).
That the British government would have readily considered doing this can hardly be surprising: by 1831, 7 million pounds per year-a full third of government expenditure (excluding debt repayment)-was going into Poor Relief, a cause of serious political tension in that tumultuous year (Goldstone 1991: 324).This was also near the highpoint of Malthusian pressure among the British elite, and the implications of the seismic demographic events playing out were anything but lost on the state's rulers.
Thus, in demographically-exploding 19 th century Britain, the state was willing to commit significant resources to encouraging emigration, settlement, and poor relief.Where else was British state spending going in these years?
This is a good time to consider the British budget as a whole in this era.In fact, viewing budgetary spending from the mid-18 th century to the 1830s is extremely revealing of some of the major developments unfolding in British society over this period.
The starkest development by far is the growth of military spending, as can be seen in Figure 11.A number of points are notable here.Firstly, military spending grew to the extent that it did for a very good reason.That is, Britain spent most of the period from the mid-18th century to 1815 more or less continuously at war.However, more significant for our purposes is what happens when, at the end of this prolonged period of war in 1815, Britain finds herself again at peace.At this point the British budget recedes-as does military spending, which makes up the majority of the budget.However, it does not return to anything close to pre-war levels, even adjusting for inflation and demographic growth.Rather, a "ratchet effect" (Tilly 2005: 130) sees spending drop, but remain at about two thirds of its wartime peak for the next several decades-even spiking briefly to near that level in the 1830s, as Figure 11   But again, there is a good reason for this.As Tilly (1985) famously pointed out, war-making is often the basis of state-making and state-building-and the British state was no exception.Tilly also describes specifically how the modern British state developed on the back of the protracted wartime expansion of these years.
In times of normal population growth, paring back a state and military swollen by years of wartime mobilization would be challenging-a major (albeit not the only) upshot being: what do all the people who are now surplus to requirements do?
But post-Waterloo Britain was not a normal society experiencing normal population growth-and a significant paring back of the state would not have been merely challenging, but potentially catastrophic-unless an ingenious way could be found to occupy the tens of thousands of people suddenly turned out of military life and public service.Thus, more than a moderate scaling back of the state and military was virtually impossible in a Britain where exploding demographics meant that finding employment for the thousands of people being added on to the population every year was an extreme challenge even without sudden shocks such as the mass standing down of military forces.
Thus, what happened after 1815 could be considered in terms of the grand design of an interventionist state, in terms of path dependence, or in terms of an equilibrium solution to the problems posed by runaway demographic growth.However, the basic fact was that after expanding the state and military to such an extent over the preceding half-century or more, there was no going back; the sheer momentum of prevailing structural forces meant that something close to that level of military and civil state mobilization would have to be maintained indefinitely.And the political and economic conjuncture of early 19 th century Britain dictated the solution to the question of how Britain would usefully direct this war-hardened state capacity.The need for lands to settle surplus population, the need for markets to absorb the exports of Britain's manufacturing industry, the need for cheap raw materials to fuel that manufacturing, and the need for something for soldiers and sailors and Oxbridge-educated military officers and administrators to do-all of this complicated by the recent loss of sovereignty over the thirteen American colonies-made a new round of imperial/colonial expansion what later generations would call a no-brainer, killing multiple birds with one stone.Thus, contrary to the impression given by the rather arcane debates surrounding attempts to "theorise" imperialism in this era, British imperialism would appear to have been over-determined by the panoply of overwhelming material forces prevailing at this time.
In case this bundle of needs and counterfactuals sounds overly abstract, let us consider what happened in France upon the defeat of Napoleon.Alongside the serious recession caused by France's military defeats and reversals, officer misgivings at chronically low pay in the country's straitened post-war circumstances fueled much of the elite discontent that would figure in subsequent revolutionary upsurges.Goldstone (1991: 306-311) points out that the servicing of debts to occupying powers led to serious implications for state finances, and to the cutting of state and military expenditure to the bone.As a result, military officer wages stagnated to the point that they ended up at half the British or Prussian rate.This discontent was compounded when the Restoration led to great resentment among military officers and other state functionaries as the meritocracy of the Napoleonic period receded, and senior officer positions were given to returning emigrés.Rather than opening up new elite routes/positions, upper elite prestige continued much as before, tied to senior offices of state and to landholding.This elite discontent contributed to continuing instability in France, highlighting the danger of what happens when a nation's officer corps is not kept happy after war-something Britain managed to avoid by maintaining her peacetime forces at a high level, and finding an economically productive outlet for them in the expansion of the Empire.Barnett (1974: 272-273, cited in Tilly 2005: 128) points out that after 1815, up to three quarters of Britain's standing infantry forces would be garrisoned overseas throughout the empire.The British Army was now for the first time a largely colonial one10 : After 1815 therefore the British army's role was very different from in the century after 1715.It was no longer needed, however intermittently, in struggles against great powers, or in deliberate imperial expansion.Nevertheless, the existing empire, irrelevant though now it seemed, had to be garrisoned and its borders protected against native unrest.The peacetime strength of the British standing army rose to the unprecedented figure of over 100,000 men.However, up to three quarters of the infantry were overseas, in the empire.The army now became for the first time a largely colonial army, whose units might be exiled for a decade or more at a time.
Colquhon, in his survey of the British Empire at the end of the Napoleonic Wars, put the number of military officers (including those on half-pay and superannuated) at 69,000 people (Tilly 2005: 115)-a substantial figure, given that according to Lindert's estimates of changes in major occupations between 1755 and 1811, the "Army" occupation as a whole had gone from 10,000 in 1755 to 266,000, or average 6% annual growth (Tilly 2005: 117).
However, this latest imperial expansion was no boondoggle; O'Rourke and Findlay (2007) demonstrate that, well-known arguments to the contrary notwithstanding, from the mercantile era onwards, the health of the British economy was tied up inextricably in securing favourable trade routes and trade relationships overseas, and this was predicated on the successful employment of strong-arm military tactics and the ability to project power on an increasingly global checkerboard.
Given the scale of the popular and elite safety valve entailed by ESCISE-not to mention its key economic or quasi-economic role in keeping 'Britain Inc.' afloat-we can only wonder at how British political stability might have fared in the absence of any one element of this nexus.

Conclusion
This article set out with two main goals, both of them guided by Goldstone's illuminating yet non-exhaustive excursus on the case of 19th century England/Britain.Those goals were, firstly, to establish with greater precision the relative political stability/instability of Britain during the period in question, using political-violence-induced fatalities as a proxy for instability-something beyond the scope of Goldstone's limited case study; and secondly, to revisit Goldstone's causal analysis of Britain's relative political stability.Accepting that the domestic economic factors Goldstone's explanation highlighted were clearly necessary to British political stability, we asked whether they were entirely sufficient.
We believe that on both of these counts, this article has cast some significant new light on the important questions raised initially by Goldstone.Thanks to our unique TCD Political Violence dataset, we believe we have been able to establish a more rigorous and systematic picture of the real levels of deadly political violence occurring in 19th century Britain than any heretofore study, with important implications for how we think about the relative political stability of the era.The picture generated by our data analysis broadly accords with the general outline of relative political violence and instability presented by Goldstone, and by other relevant studies such as that of Tilly (2005), but brings considerable clarity to the area.Simply put, this picture is of a Britain increasingly free of serious political violence and political instability over the course of the 19th century, albeit punctuated by a dramatic uptick in both in the early 1830s, and by more minor paroxysms of violence and instability at other times.The great exception here, of course, is Ireland-part of the 'British state' in the sense of the United Kingdom, but not part of Britain as such-where political violence and instability were consistently much higher.
As for the other dimension to this exercise-that concerning the causal story behind these patterns of relative political violence and political stability-we believe that we have highlighted some important causal factors underemphasised in other accounts of the epoch.While we fully accept the crucial nature of the domestic economic circumstances underlined by Goldstone-for instance the employment opportunities generating by expanding industry and an expanding civil service-we suggest that these could not on their own have been responsible for the maintenance of political stability in 19th century Britain.We point to the nexus of emigration, settler-colonialism, and imperial state expansion (ESCISE) that was so intertwined in British life in this period, and we suggest that without any one of these 'safety valves', it would be difficult to imagine 19th century Britain remaining quite so stable as it did.
Needless to say, it is not news that Britain's imperial state expanded in the 19th century, nor that significant numbers left Britain and settled in other lands.But while these facts are not entirely unknown, we argue that they are obscurely known, and poorly articulated together-let alone to the dynamics of relative political stability to be seen within Britain herself.
Compared to most other European countries, Britain was positively hemorrhaging people from its shores in the 19th century-mostly to the United States.And yet, while the twee idea of the 'WASP' still has some resonance in the ethnic imaginary of the US, the country's truly most numerous 'ethnic diasporas'-'British-Americans' or 'English-Americans', depending on one's perspective-barely register as categories.Americans rarely self-identify as such, even when invited to do so.According to the US Census Bureau's 2009 American Community Survey, 13% of the US population reported one or other type of British ancestry-a wild undercount by any estimate.To the extent that people do imagine an American majority tracing back roots to Britain, vague ideas of Mayflower pilgrims are conjured up.But most Americans of British heritage are not seeded from the Mayflower pilgrims, or even from the relative glut of 18th century British pioneers-they are much more likely to be descended from the unprecedented numbers of British immigrants flowing into the country from the 1820s onwards.
That this great migration-coinciding with the fanning out of European settlers across the American continent-took place after American independence tells us that we need to rethink our ideas about British settler-colonialism in the Americas.That is, the loss of British sovereignty over the Thirteen Colonies did not bring a stop to British settler-colonialism in the proto-US.The shucking off of the British state's geopolitically-motivated constraints on further westward expansion freed this process up.Settler-colonialism continued and expanded, largely at the hands of British people or early-generation British diaspora-it was simply no longer under the control of the British state.
We suggest that, while at times enlightening, academic debates on how best to theorise colonialism and imperialism have often taken esoteric directions that have not always been analytically useful.We view British imperialism and colonialism differently, as over-determined by the sheer momentum of a panoply of almost irresistible material and structural forces.These dynamics penetrated British state and society, and are inseparable from the domestic patterns of political stability and instability considered in this article.The anomaly of Ireland is an interesting case in point: while Ireland was a de jure part of the United Kingdom, Ireland's high levels of political violence were driven by the vicissitudes of its de facto colonial relationship with Britain.And while France still struggled with serious subsistence crises throughout the 19th century, Britain benefited from Ireland's captive food imports-again in keeping with the ESCISE-based pattern of British political stability maintenance.
All of these findings naturally open up other questions.To what extent was the ESCISE nexus we speak of conscious British state policy?To what extent were ruling elites conscious of the relationship between the ESCISE nexus and the maintenance of domestic political stability?And why did France not follow suit?If it is true that France was suffering greater relative structural demographic strain in this era (albeit with slower population growth than Britain), why was the French emigration rate among the lowest in Europe, and outstripped by British emigration by a factor of fifty for much of the 19th century?And why, if structural demographic pressures were so intense there, was France experiencing significant in-migration from countries such as Belgium and Italy in this period?These questions require further research, and suggest deeper case studies utilising more qualitative and mixed methods.

Figure 9 .
Figure 9. Experimental results: estimated frequency of reports of political violence events with fatalities by decade, 1785-1900.
clearly shows.

Table 3 .
Experimental Results: Estimated Frequency of Reports of Political Violence Events by decade, 1785-1900

Table 4 .
Gold Standard Analysis of Reports from 1831