We’re really interested in how we ended up living in a democracy - and we think that you should be too. Not only is it just a surprising outcome that’s worth exploring on its own, but also having some idea of which factors lead to countries becoming more democratic and staying that way allows us to be more conscious of democratic backsliding. This is something that we’re especially conscious of, which might become more prominent in the age of advanced AI. But in this post, we’re thinking purely about democracy - through case studies, looking at existing literature on the topic and our own hypothesising, and present strong civil society in particular, alongside strong rule of law and institutionalised political parties, as factors that raise levels of democracy and sustain them, as well as presenting some ideas for why they might be much stronger now than in previous centuries.
If you’re reading this right now, the chances are that you’re living in a democracy. Most importantly, you probably view democracy as a standard political system: one that may be imperfect or the least worst option, but not particularly remarkable or strange.
But historically, this is very strange. Until the mid 19th century, almost no one was living in a democracy, and today, billions of people are across roughly half of the world's countries[1]. Elections spread during the 19th century, but this often took countries to a state of ‘electoral autocracy ’- where elections were held, but they didn't reach democratic standards of freedom and fairness. After this period, the 20th-century trend was towards electoral and liberal democracy (an electoral democracy is a regime where citizens have protected rights, are equal before the law, and checks and balances on the power of ruling officials are present).
There’s also no certainty that democracies are here to stay - in fact, some have argued that we’re living in a brief political oasis: one where a selection of factors are balanced in a delicate equilibrium, resulting in the empowerment to choose our own leaders. This graph already shows a worrying decline over roughly the past 15 years: perhaps the 2010s were the global democratic peak, and it’s only going to get worse from here.
All of this has made my research group curious, so we’ve tried to develop a better understanding of the factors at play that have gotten us to the democracies we’re in now, and especially how these are likely to be changed by advanced AI now and in the coming years.
So, in this blog post, we are presenting the results of a literature review, some useful case studies and our own hypothesising on the factors that affect democracies (whether they increase, decrease or sustain existing levels of democracy), where we hope to make the insights of experienced researchers more accessible to the EA and AI safety communities. We are by no means experts, and this work hasn’t completely cleared up our own confusion about how democracies have come about and remained, but it’s certainly been an educational first step towards that goal, and we hope that this might be the case for others too.
Three clusters to consider
When looking at how democracies arose, we need to consider the previous state a country was in. One key distinction to consider is whether a country (1) experienced a settler-led population reconstitution (2) experienced externally driven regime change after military defeat, or (3) shifted from a previous equilibrium towards becoming more democratic, through primarily domestic dynamics, rather than as a direct result of external shocks.
Complete or near-complete restructuring of the power dynamics within a country due to colonisation and/or settlement by another group.
For example, the large-scale settlement by Europeans in what later became the US, causing a near-complete restructuring of power relations as settler institutions became dominant.
Australia and New Zealand are somewhat similar examples where the introduction of private landownership widely redistributed economic power.
A country facing devastation after losing a war, and being restructured by the winning powers.
For example, democracy in Western Germany after WWII was significantly shaped by the UK, France, and the US, which were already democracies.
Shift from internal pressure.
This includes the UK, France, and many more to some extent. See the Case Studies section for an in-depth explanation of the onset of democracy.
These are the cases where there is no obvious external factor to point to; where democratic transition sprung from a previous autocratic period. So, while we think that all three cases contain interesting dynamics, we will focus on this case.
In particular, we are more interested in the advent of early democracies, because they might point more cleanly to underlying causes of democracies rather than outside influence or selection effects[2]. So to give some context to our theorising, we’ll present three very short histories of democratic development in three different countries - France, England and Japan.
Three very short histories of democratic development
France
James Eugene Farmer, Versailles and the Court under Louis XIV (1905), via Wikimedia Commons
The 1789 Revolution was triggered by state bankruptcy (due to high-interest loans that were used to finance the American war), which forced Louis XVI to convene les Etats-Généraux, which he couldn't control once assembled. A crack in elite coalition support, with the liberal bourgeoisie and much of the nobility withdrawing cooperation from the Crown, created the opening. Extreme poverty, following the disastrous 1788 harvest failure and harsh winter, caused people to revolt, which radicalised the process and pressured the revolutionary assemblies into abolishing the privileges and legal status of the governing nobility.
But what came out of the French Revolution wasn't a stable democracy: it was a cycle of republics, empires, and restorations, because none of the structural underpinnings for consolidation were in place.
In fact, democratic participation was barely featured from 1799 until the Revolution of 1848, which was caused by dire economic conditions stemming from poor harvests and business failures, and pressure from the French people holding large banquets across France for seven months to increase voting rights, circumventing a ban on such political meetings. The highly unpopular King Louis-Philippe failed to suppress this revolution, leading to his abdication and the formation of the Second Republic (1848–52), which featured France's first presidential elections with universal male suffrage.
The president Louis-Napoleon managed to establish himself as Emperor Napoleon III towards the end of his presidential term, although he maintained broad public support, repeatedly ratified by plebiscite, for most of his rule until 1870.
Democratic stability only came with the Third Republic (1870), which, to a significant extent, managed to be established only because opponents of the democratic approach didn't manage to coordinate around what alternative government should be created.
The recurring pattern throughout this history is that economic distress and suffering of the population, together with political failure by the regime and its supporters, lead to changes in the political order, often through revolutions. Improved institutions and practice in elections likely helped the democratic regime to be sustained later where it failed earlier.
England
Arthur C. Michael, King John signing the Magna Carta reluctantly, before 1945, via Wikimedia Commons
England’s democratic development actually took place well before the Industrial Revolution. This can be traced back to the early 13th century, when forty barons rebelled against King John over excessive taxation, disastrous war defeats such as the loss of Normandy, arbitrary imprisonment, and violations of property rights. They captured London and forced the King of England to adopt the famous Magna Carta in 1215.
While the Magna Carta didn’t actually create parliament, it did strengthen the idea that the King couldn’t raise taxes autonomously. This developed further in 1265, when a parliament was brought together, which comprised nobles and clergy, as well as knights from the shires and representatives from the towns. This was then reinforced by Edward I through what’s known as the Model Parliament of 1295, which established the idea that taxation required pretty broad political consent, which also set a precedent for future English parliaments.
The following centuries were still turbulent. The Personal Rule from 1629-1640 saw Charles I govern without parliament, intensifying the conflict between royal prerogative and parliamentary consent. This erupted into civil war and culminated in the Glorious Revolution of 1688, when leading political figures invited William of Orange to invade the country. This ultimately led to Parliament passing the Bill of Rights, which strengthened Parliament’s position by affirming parliamentary sovereignty, free elections, freedom of expression within Parliament, and the prohibition of a standing army without Parliament's consent.
The next two centuries brought enormous wealth to the country with the UK's colonies and the Industrial Revolution. The latter specifically led to various revolts, most notably agitation over rotten boroughs, which were essentially new cities that emerged with the Industrial Revolution but lacked political representation. More than four decades of violent social unrest for political rights led to a succession of reforms, ultimately enfranchising 60% of working-class men by 1884.
The First World War had a decisive effect on suffrage, leading to the extension of the vote to all men over 21 and some women over 30. Before the war, the right to vote depended on being a resident of the same place for 12 consecutive months, and it could have been politically disastrous to continue to deny the vote of men who were away to fight and produce weapons. For women, the picture was more nuanced. While it is commonly held that women were ‘rewarded’ for their service, in practice this is not the case. Only women over 30 who were long-term residents or owners (or the wives of men who were) could finally vote, even though most of the women who fought and supported the war effort were under 30. In fact, most historians agree that the suffragettes' longstanding efforts were the main variable. And even then, MPs made sure to only enfranchise a small share of women so they would not outnumber men. This situation eventually changed in 1928, when the UK finally achieved equal voting rights, thereby fulfilling one of the core requirements of democracy.
We can pick out some key underlying factors relevant to the gradual development of democracy here - in particular, the successful coordination of groups opposed to King John, the diffusion of power from one leader to a broader range of representatives, as well as factors including economic growth and successful resistance movements. There are likely other more distal factors at play too, including urbanisation and higher levels of education, that aren’t captured in this description.
Japan
Goseda Hōryū II, Conference on Drafting a Constitution (1926), Meiji Memorial Picture Gallery, via Wikimedia Commons
Emperors and clans ruled for most of Japan’s history, and society was organised into a rigid hierarchy in which samurai, farmers, and merchants each had distinct rights. This changed when the Meiji oligarchs declared a constitution in 1889, creating a two-chamber parliament with a sovereign and a representative lower house. However, ultimate authority remained with the emperor, as had previously been the case.
The constitution came into being partly to revise unequal trade agreements with Western powers, which did not regard Japan as ‘civilised’ because it lacked the rule of law and a constitution. It was also shaped by growing popular domestic movements that called for greater representation, more political rights, and constitutional government.
To specifically choose this constitutional model, the Meiji leadership sent Itō Hirobumi, an influential Japanese statesman, to Europe to study its different political systems. He remained unconvinced by the British and French ones, but was more attracted to the Prussian system, and thus pushed his government to adopt it into the constitution. Later, between 1912 and 1932, political parties began to gain real influence. In essence, the Meiji oligarchs in power were losing legitimacy and had no real successors. The Meiji Constitution also left important questions about government formation unresolved. Because it did not clearly state how cabinets should be chosen or whether they needed majority support in the elected lower house, party-based government was able to emerge gradually as a political convention rather than as a formal constitutional requirement.
In the end, after these parties started to be much more powerful, they declared the 1925 Universal Manhood Suffrage Act, which expanded the voting pool from 3 million men to 15 million.
Ultimately, examining the broader context of how democracy evolved in Japan reveals that various factors came into play. The willingness to be looked up to by foreign powers, growing internal pressures for more rights by the population, and later on, the lack of legitimate descendants proved critical in changing political systems.
Some underlying factors
We’ve now examined three case studies, that fall into the ‘third cluster’; i.e. where no obvious external factor triggered the democratisation of the state. So this brings us to the question of what underlying factors might influence levels of democracy, without assuming that democracy must already exist elsewhere.
The most helpful and evidence-driven source we could find for this was the book Why Democracies Develop and Decline, edited by Michael Coppedge, Amanda B. Edgell, Carl Henrik Knutsen, and Staffan I. Lindberg. This book is based on findings from the V-Dem (Varieties of Democracy) dataset, the world’s largest, most detailed database on democracy, measuring hundreds of political and social indicators across over 200 countries and territories from 1789 to the present day. Rather than viewing democracy as a simple "yes/no" state (like just checking if a country holds elections), V-Dem breaks it down into multiple measurable dimensions, using over 600 specific indicators rated by a global network of more than 4,200 country experts. Countries may perform well on some indicators, and worse on others - for example, a country might have universal suffrage and elected officials but a heavily censored media, and these are combined to give an overall democracy score in a way that rewards countries for doing well across the board, but penalises them when one essential democratic component is missing.
Here, we’ll present three factors that feature in this analysis, which comprise what the authors call the ‘protective belt’ of democracy, that together form part of the picture of how democracies arise, survive and stabilise: civil society, the rule of law and institutionalised political parties[3]. These are three of the V-Dem indicators/indices that are most strongly linked to higher and more durable levels of democracy, and in the case of civil society, a higher likelihood of democratic upturns. They are also interlinked: when one is strong, it strengthens the others, but the inverse also readily applies.
Civil Society
This refers to the capacity of a society to organise itself and exert strength: through means like protest, strike action, labour unions, forming community and charitable organisations and social movements - and it can trigger sudden increases in levels of democracy in the case of non-violent resistance campaigns, and sustainment of existing levels of democracy through other means.
A stronger civil society can be measured by country experts through the indicator of how involved ordinary people are in civil society organisations, which have to be legitimately voluntary organisations that aren’t state-controlled, and ideally, there should be lots of them to choose from.
It is strongly associated with higher levels of what’s known as procedural democracy, which is about whether rulers are chosen through real, competitive elections with broad participation. It is also associated with positive changes in democracy in a country year-on-year, so countries with high civil society measures are likely to experience more democratising movements than countries with lower levels, and it can also guard against autocratisation. This means that strong civil society promotes democratic upturns, and also makes democratic decline less likely or less severe.
One key form this can take is non-violent resistance campaigns, which include marches, protests, strikes and boycotts. They can be crucial proximal factors (i.e. immediate triggers) that can turn long-standing authoritarian stability into a sudden democratic opening. The reason that we specify non-violent campaigns is that they are particularly strongly associated with upturns in democracy, including continued upturns in the years after the movement. This can be understood in the sense that non-violent campaigning applies political pressure to the existing regime, but without providing any remotely supportable excuse for violent repression in return (this last bit is our own hypothesising)
Rule of Law
This is the second of the three features that make up the protective belt of democracy, that together help democracy survive and stabilise once democratic openings occur. This makes sense: when the rule of law is strong, ruling officials have to follow the law just like everyone else, which limits the possibility of corruption, power grabs or unconstitutional actions overall.
It also means in a liberal democracy that opposition parties can compete without being threatened by targeted legal harassment, and civil society groups can operate without fear of police raids, politicised prosecutions or outright bans.
However, the rule of law can still be applied to a weaker extent in non-democracies - for example, in predictable court procedures for ordinary disputes, consistent contract enforcement or limits on arbitrary behaviour by lower officials. These partial legal constraints do not by themselves cause democratisation, but they make it harder for rulers to crush opposition arbitrarily once pressure for reform emerges, if legal processes take place at least on some level - which could be what buys the time for a reform movement to take hold.
Institutionalised Political Parties
Strong political parties primarily help sustain levels of democracy in several ways: by providing a channel through which people can express their views and demands, by holding the other parties (especially the ruling party) to account, and by safeguarding against a political vacuum: which might provide the perfect opportunity for the military or any autocratic hopeful to take charge. This also needs to be seen separately from measures of civil society, because civil society can mobilise pressure, but parties translate pressure into electoral competition and government formation.
It's possible for institutionalised political parties to exist in autocracies as well as democracies - it's just that they don't compete totally freely for power in this case. To illustrate this, in an electoral autocracy, there may be gerrymandering or media suppression that makes non-incumbents less likely to be elected. For a party to be institutionalised it needs to be stable, organised and internally disciplined to some extent, and to have embedded roots in society - for example through local branches, youth wings, and membership opportunities.
So why are these factors stronger now than centuries ago, in a way that has made more democracies pop up?
The case studies show how tumultuous the process of democratic evolution can be, but they might also unveil common denominators that might help explain why civil society, rule of law and parties are stronger now than centuries ago.
We will now put forward some of our extremely rough ideas on why civil society in particular might have strengthened in roughly the last 200 years, as this is the factor that is shown to drive specifically upturns in democracy most clearly.
Industrialisation and urbanisation have brought us to a point in history that is pretty unique, where populations have been living more closely together within cities - allowing workers, students, professionals, religious groups, newspapers, trade unions, charities, campaign groups, and political parties to form dense networks. These networks may create a stronger civil society because people can meet, share ideas, protest, strike, and coordinate demands for rights.
Workers are generally more productive when they’re healthy and educated(at least to some minimum threshold, e.g. literacy), so in an environment where businesses exist, with business owners wanting to maximise productivity and having resources available to set up hospitals and schools, it makes sense that a more healthy and educated population (as compared to earlier societal structures) is what industrialised societies eventually converge towards. However, unluckily for autocrats, if you’re educated and not on the verge of death, you’re probably more likely to be able to engage in civil society to promote your own political rights.
Technological developments in this period introduced the concept of free time - which gives you more time to organise outside of the state.
We also have some other factors in mind that aren’t directly related to civil society, but are probably still important. For example, as the state’s capacity increases (which is in the interest of improved productivity), it becomes harder to rule with fewer people, as state systems simply become too complex. This could also be the case for state militaries in general over the last few centuries: militaries have become more powerful and this has been facilitated in part by vast increases in numbers of service personnel, most of whom were ordinary citizens rather than elite warriors, who may have been more sympathetic to policies that benefitted the population, and less inclined to support rulers unconditionally.
Conclusion
At this point, it’s worth reiterating that this article does not attempt to get anywhere near a complete solution, but instead presents a reasonable model that might be useful to investigate how other factors could affect levels of democracy.
The overall message that we want to get across is that democracy can only really come about and remain in place as the result of the delicate interplay of social organisation, legal constraint and well-entrenched political competition. To take democracy from a fad to an enduring political reality, these factors are what we must work to protect.
There are many more factors that are explored in the book, including very interesting distal factors like climate or distance from natural harbours. But we’re interested in trying to understand how levels of democracy change, so we won’t be considering these practically constant factors.
We’re really interested in how we ended up living in a democracy - and we think that you should be too. Not only is it just a surprising outcome that’s worth exploring on its own, but also having some idea of which factors lead to countries becoming more democratic and staying that way allows us to be more conscious of democratic backsliding. This is something that we’re especially conscious of, which might become more prominent in the age of advanced AI. But in this post, we’re thinking purely about democracy - through case studies, looking at existing literature on the topic and our own hypothesising, and present strong civil society in particular, alongside strong rule of law and institutionalised political parties, as factors that raise levels of democracy and sustain them, as well as presenting some ideas for why they might be much stronger now than in previous centuries.
If you’re reading this right now, the chances are that you’re living in a democracy. Most importantly, you probably view democracy as a standard political system: one that may be imperfect or the least worst option, but not particularly remarkable or strange.
But historically, this is very strange. Until the mid 19th century, almost no one was living in a democracy, and today, billions of people are across roughly half of the world's countries[1]. Elections spread during the 19th century, but this often took countries to a state of ‘electoral autocracy ’- where elections were held, but they didn't reach democratic standards of freedom and fairness. After this period, the 20th-century trend was towards electoral and liberal democracy (an electoral democracy is a regime where citizens have protected rights, are equal before the law, and checks and balances on the power of ruling officials are present).
This is already surprising, given that this development was happening in the same two centuries in which states became vastly more capable at surveillance, propaganda, border control, and military logistics.
There’s also no certainty that democracies are here to stay - in fact, some have argued that we’re living in a brief political oasis: one where a selection of factors are balanced in a delicate equilibrium, resulting in the empowerment to choose our own leaders. This graph already shows a worrying decline over roughly the past 15 years: perhaps the 2010s were the global democratic peak, and it’s only going to get worse from here.
All of this has made my research group curious, so we’ve tried to develop a better understanding of the factors at play that have gotten us to the democracies we’re in now, and especially how these are likely to be changed by advanced AI now and in the coming years.
So, in this blog post, we are presenting the results of a literature review, some useful case studies and our own hypothesising on the factors that affect democracies (whether they increase, decrease or sustain existing levels of democracy), where we hope to make the insights of experienced researchers more accessible to the EA and AI safety communities. We are by no means experts, and this work hasn’t completely cleared up our own confusion about how democracies have come about and remained, but it’s certainly been an educational first step towards that goal, and we hope that this might be the case for others too.
Three clusters to consider
When looking at how democracies arose, we need to consider the previous state a country was in. One key distinction to consider is whether a country (1) experienced a settler-led population reconstitution (2) experienced externally driven regime change after military defeat, or (3) shifted from a previous equilibrium towards becoming more democratic, through primarily domestic dynamics, rather than as a direct result of external shocks.
In particular, we are more interested in the advent of early democracies, because they might point more cleanly to underlying causes of democracies rather than outside influence or selection effects[2]. So to give some context to our theorising, we’ll present three very short histories of democratic development in three different countries - France, England and Japan.
Three very short histories of democratic development
France
James Eugene Farmer, Versailles and the Court under Louis XIV (1905), via Wikimedia Commons
The 1789 Revolution was triggered by state bankruptcy (due to high-interest loans that were used to finance the American war), which forced Louis XVI to convene les Etats-Généraux, which he couldn't control once assembled. A crack in elite coalition support, with the liberal bourgeoisie and much of the nobility withdrawing cooperation from the Crown, created the opening. Extreme poverty, following the disastrous 1788 harvest failure and harsh winter, caused people to revolt, which radicalised the process and pressured the revolutionary assemblies into abolishing the privileges and legal status of the governing nobility.
But what came out of the French Revolution wasn't a stable democracy: it was a cycle of republics, empires, and restorations, because none of the structural underpinnings for consolidation were in place.
In fact, democratic participation was barely featured from 1799 until the Revolution of 1848, which was caused by dire economic conditions stemming from poor harvests and business failures, and pressure from the French people holding large banquets across France for seven months to increase voting rights, circumventing a ban on such political meetings. The highly unpopular King Louis-Philippe failed to suppress this revolution, leading to his abdication and the formation of the Second Republic (1848–52), which featured France's first presidential elections with universal male suffrage.
The president Louis-Napoleon managed to establish himself as Emperor Napoleon III towards the end of his presidential term, although he maintained broad public support, repeatedly ratified by plebiscite, for most of his rule until 1870.
Democratic stability only came with the Third Republic (1870), which, to a significant extent, managed to be established only because opponents of the democratic approach didn't manage to coordinate around what alternative government should be created.
The recurring pattern throughout this history is that economic distress and suffering of the population, together with political failure by the regime and its supporters, lead to changes in the political order, often through revolutions. Improved institutions and practice in elections likely helped the democratic regime to be sustained later where it failed earlier.
England
Arthur C. Michael, King John signing the Magna Carta reluctantly, before 1945, via Wikimedia Commons
England’s democratic development actually took place well before the Industrial Revolution. This can be traced back to the early 13th century, when forty barons rebelled against King John over excessive taxation, disastrous war defeats such as the loss of Normandy, arbitrary imprisonment, and violations of property rights. They captured London and forced the King of England to adopt the famous Magna Carta in 1215.
While the Magna Carta didn’t actually create parliament, it did strengthen the idea that the King couldn’t raise taxes autonomously. This developed further in 1265, when a parliament was brought together, which comprised nobles and clergy, as well as knights from the shires and representatives from the towns. This was then reinforced by Edward I through what’s known as the Model Parliament of 1295, which established the idea that taxation required pretty broad political consent, which also set a precedent for future English parliaments.
The following centuries were still turbulent. The Personal Rule from 1629-1640 saw Charles I govern without parliament, intensifying the conflict between royal prerogative and parliamentary consent. This erupted into civil war and culminated in the Glorious Revolution of 1688, when leading political figures invited William of Orange to invade the country. This ultimately led to Parliament passing the Bill of Rights, which strengthened Parliament’s position by affirming parliamentary sovereignty, free elections, freedom of expression within Parliament, and the prohibition of a standing army without Parliament's consent.
The next two centuries brought enormous wealth to the country with the UK's colonies and the Industrial Revolution. The latter specifically led to various revolts, most notably agitation over rotten boroughs, which were essentially new cities that emerged with the Industrial Revolution but lacked political representation. More than four decades of violent social unrest for political rights led to a succession of reforms, ultimately enfranchising 60% of working-class men by 1884.
The First World War had a decisive effect on suffrage, leading to the extension of the vote to all men over 21 and some women over 30. Before the war, the right to vote depended on being a resident of the same place for 12 consecutive months, and it could have been politically disastrous to continue to deny the vote of men who were away to fight and produce weapons. For women, the picture was more nuanced. While it is commonly held that women were ‘rewarded’ for their service, in practice this is not the case. Only women over 30 who were long-term residents or owners (or the wives of men who were) could finally vote, even though most of the women who fought and supported the war effort were under 30. In fact, most historians agree that the suffragettes' longstanding efforts were the main variable. And even then, MPs made sure to only enfranchise a small share of women so they would not outnumber men. This situation eventually changed in 1928, when the UK finally achieved equal voting rights, thereby fulfilling one of the core requirements of democracy.
We can pick out some key underlying factors relevant to the gradual development of democracy here - in particular, the successful coordination of groups opposed to King John, the diffusion of power from one leader to a broader range of representatives, as well as factors including economic growth and successful resistance movements. There are likely other more distal factors at play too, including urbanisation and higher levels of education, that aren’t captured in this description.
Japan
Goseda Hōryū II, Conference on Drafting a Constitution (1926), Meiji Memorial Picture Gallery, via Wikimedia Commons
Emperors and clans ruled for most of Japan’s history, and society was organised into a rigid hierarchy in which samurai, farmers, and merchants each had distinct rights. This changed when the Meiji oligarchs declared a constitution in 1889, creating a two-chamber parliament with a sovereign and a representative lower house. However, ultimate authority remained with the emperor, as had previously been the case.
The constitution came into being partly to revise unequal trade agreements with Western powers, which did not regard Japan as ‘civilised’ because it lacked the rule of law and a constitution. It was also shaped by growing popular domestic movements that called for greater representation, more political rights, and constitutional government.
To specifically choose this constitutional model, the Meiji leadership sent Itō Hirobumi, an influential Japanese statesman, to Europe to study its different political systems. He remained unconvinced by the British and French ones, but was more attracted to the Prussian system, and thus pushed his government to adopt it into the constitution. Later, between 1912 and 1932, political parties began to gain real influence. In essence, the Meiji oligarchs in power were losing legitimacy and had no real successors. The Meiji Constitution also left important questions about government formation unresolved. Because it did not clearly state how cabinets should be chosen or whether they needed majority support in the elected lower house, party-based government was able to emerge gradually as a political convention rather than as a formal constitutional requirement.
In the end, after these parties started to be much more powerful, they declared the 1925 Universal Manhood Suffrage Act, which expanded the voting pool from 3 million men to 15 million.
Ultimately, examining the broader context of how democracy evolved in Japan reveals that various factors came into play. The willingness to be looked up to by foreign powers, growing internal pressures for more rights by the population, and later on, the lack of legitimate descendants proved critical in changing political systems.
Some underlying factors
We’ve now examined three case studies, that fall into the ‘third cluster’; i.e. where no obvious external factor triggered the democratisation of the state. So this brings us to the question of what underlying factors might influence levels of democracy, without assuming that democracy must already exist elsewhere.
The most helpful and evidence-driven source we could find for this was the book Why Democracies Develop and Decline, edited by Michael Coppedge, Amanda B. Edgell, Carl Henrik Knutsen, and Staffan I. Lindberg. This book is based on findings from the V-Dem (Varieties of Democracy) dataset, the world’s largest, most detailed database on democracy, measuring hundreds of political and social indicators across over 200 countries and territories from 1789 to the present day. Rather than viewing democracy as a simple "yes/no" state (like just checking if a country holds elections), V-Dem breaks it down into multiple measurable dimensions, using over 600 specific indicators rated by a global network of more than 4,200 country experts. Countries may perform well on some indicators, and worse on others - for example, a country might have universal suffrage and elected officials but a heavily censored media, and these are combined to give an overall democracy score in a way that rewards countries for doing well across the board, but penalises them when one essential democratic component is missing.
Here, we’ll present three factors that feature in this analysis, which comprise what the authors call the ‘protective belt’ of democracy, that together form part of the picture of how democracies arise, survive and stabilise: civil society, the rule of law and institutionalised political parties[3]. These are three of the V-Dem indicators/indices that are most strongly linked to higher and more durable levels of democracy, and in the case of civil society, a higher likelihood of democratic upturns. They are also interlinked: when one is strong, it strengthens the others, but the inverse also readily applies.
Civil Society
This refers to the capacity of a society to organise itself and exert strength: through means like protest, strike action, labour unions, forming community and charitable organisations and social movements - and it can trigger sudden increases in levels of democracy in the case of non-violent resistance campaigns, and sustainment of existing levels of democracy through other means.
A stronger civil society can be measured by country experts through the indicator of how involved ordinary people are in civil society organisations, which have to be legitimately voluntary organisations that aren’t state-controlled, and ideally, there should be lots of them to choose from.
It is strongly associated with higher levels of what’s known as procedural democracy, which is about whether rulers are chosen through real, competitive elections with broad participation. It is also associated with positive changes in democracy in a country year-on-year, so countries with high civil society measures are likely to experience more democratising movements than countries with lower levels, and it can also guard against autocratisation. This means that strong civil society promotes democratic upturns, and also makes democratic decline less likely or less severe.
One key form this can take is non-violent resistance campaigns, which include marches, protests, strikes and boycotts. They can be crucial proximal factors (i.e. immediate triggers) that can turn long-standing authoritarian stability into a sudden democratic opening. The reason that we specify non-violent campaigns is that they are particularly strongly associated with upturns in democracy, including continued upturns in the years after the movement. This can be understood in the sense that non-violent campaigning applies political pressure to the existing regime, but without providing any remotely supportable excuse for violent repression in return (this last bit is our own hypothesising)
Rule of Law
This is the second of the three features that make up the protective belt of democracy, that together help democracy survive and stabilise once democratic openings occur. This makes sense: when the rule of law is strong, ruling officials have to follow the law just like everyone else, which limits the possibility of corruption, power grabs or unconstitutional actions overall.
It also means in a liberal democracy that opposition parties can compete without being threatened by targeted legal harassment, and civil society groups can operate without fear of police raids, politicised prosecutions or outright bans.
However, the rule of law can still be applied to a weaker extent in non-democracies - for example, in predictable court procedures for ordinary disputes, consistent contract enforcement or limits on arbitrary behaviour by lower officials. These partial legal constraints do not by themselves cause democratisation, but they make it harder for rulers to crush opposition arbitrarily once pressure for reform emerges, if legal processes take place at least on some level - which could be what buys the time for a reform movement to take hold.
Institutionalised Political Parties
Strong political parties primarily help sustain levels of democracy in several ways: by providing a channel through which people can express their views and demands, by holding the other parties (especially the ruling party) to account, and by safeguarding against a political vacuum: which might provide the perfect opportunity for the military or any autocratic hopeful to take charge. This also needs to be seen separately from measures of civil society, because civil society can mobilise pressure, but parties translate pressure into electoral competition and government formation.
It's possible for institutionalised political parties to exist in autocracies as well as democracies - it's just that they don't compete totally freely for power in this case. To illustrate this, in an electoral autocracy, there may be gerrymandering or media suppression that makes non-incumbents less likely to be elected. For a party to be institutionalised it needs to be stable, organised and internally disciplined to some extent, and to have embedded roots in society - for example through local branches, youth wings, and membership opportunities.
So why are these factors stronger now than centuries ago, in a way that has made more democracies pop up?
The case studies show how tumultuous the process of democratic evolution can be, but they might also unveil common denominators that might help explain why civil society, rule of law and parties are stronger now than centuries ago.
We will now put forward some of our extremely rough ideas on why civil society in particular might have strengthened in roughly the last 200 years, as this is the factor that is shown to drive specifically upturns in democracy most clearly.
We also have some other factors in mind that aren’t directly related to civil society, but are probably still important. For example, as the state’s capacity increases (which is in the interest of improved productivity), it becomes harder to rule with fewer people, as state systems simply become too complex. This could also be the case for state militaries in general over the last few centuries: militaries have become more powerful and this has been facilitated in part by vast increases in numbers of service personnel, most of whom were ordinary citizens rather than elite warriors, who may have been more sympathetic to policies that benefitted the population, and less inclined to support rulers unconditionally.
Conclusion
At this point, it’s worth reiterating that this article does not attempt to get anywhere near a complete solution, but instead presents a reasonable model that might be useful to investigate how other factors could affect levels of democracy.
The overall message that we want to get across is that democracy can only really come about and remain in place as the result of the delicate interplay of social organisation, legal constraint and well-entrenched political competition. To take democracy from a fad to an enduring political reality, these factors are what we must work to protect.
When we talk about democracy, this is how we’re thinking about it: democracy can be measured along a variety of dimensions, but electoral democracy is the core requirement, which concerns the extent to which political leaders are elected under comprehensive voting rights in free and fair elections, and freedoms of association and expression are guaranteed.
Selection effects in the sense that democratic countries might (economically) outperform nondemocratic ones and thus survive better.
There are many more factors that are explored in the book, including very interesting distal factors like climate or distance from natural harbours. But we’re interested in trying to understand how levels of democracy change, so we won’t be considering these practically constant factors.