Why did the English adopt bird watching societies and democracy?
Exploring why some countries have more clubs and associations than others, whether this leads to democracy, and what else contributes to liberalism.
I’m currently reading Watching the English, an academic book masquerading as an airport book, by British anthropologist Kate Fox. In The Guardian, commentator Catherine Bennet was quite taken in by this appearance, and promptly came away disappointed when the book was published in 2004. I’m currently two-thirds through the 2014 update.
I’ve found it insightful and accurate. Rather than “reducing” English behaviour, as Bennett wrote, I think the book draws out key parts of the ‘English character’ in a way that helps us see the forest for the trees. Fox’s emphasis on Englishness testifies to her academic bent: the characteristics she observes are common across Great Britain, but not wanting to be sensationalist or too general, she focuses on the English, acknowledging that Scots and the Welsh have their own further idiosyncrasies.
Fox’s main thesis is that the English are particularly socially inhibited, privacy-conscious, and relentlessly moderate. One of my regrets is that she spends more time exploring ‘what’ and not enough time exploring ‘why’, though she frequently draws comparisons with Japan, where she identifies similar societal traits and common potential causes, noting that both countries, for example, are island nations isolated from far larger continents.
One particular discussion focuses on the English love of clubs and societies. It’s an interesting contrast for a book that says one of the definitive traits of an English person is a love of privacy. What made this interesting to me, with my political scientist hat on, is that a love of clubs and societies (ie: associations) has been promoted as a key cause of democracy since the days of de Tocqueville in the 1830s. The idea is that forming associations creates a strong civil society and trust between your citizens. In turn, that prompts them to push for democracy. Fox puts the English love of clubs down to what she calls a “social ‘dis-ease’”:
💬 "There is an apparent contradiction ... between our obsession with privacy and our 'clubbability'. Jeremy Paxman notes (in his 1999 book, 'The English: A Portrait of a People') that the English have a club for almost everything: 'to go fishing, support football teams, play cards, arrange flowers...'
He seems to accept de Tocqueville's pragmatic, economic explanation, that the English historically have always formed associations in order to pool resources, when they could not get what they wanted by individual effort.
I would argue that clubs are more about social needs than practical or economic ones. The English are not keen on random, unstructured, spontaneous, street-corner sociability ... We prefer to socialise in an organised, orderly manner.
Above all, as with sports and games, we need to pretend that the activity of the club or society is the real point of the gathering, and that social bonding is just a secondary side-effect."
From Fox, we have small/isolated society ➟ social ‘dis-ease’ ➟ associations. Perhaps it’s no coincidence that ‘associational’ sounds so similar to ‘asocial’. Drawing comparisons with Japan, Fox also speculates that being an island society may be a cause of the so-called English ‘dis-ease’.
Apparently unintentionally, Fox finds herself aligned with an established argument in political economy: small countries do well. Peter Katzenstein wrote the totemic Small States in World Markets in 1985, and it’s still on required reading lists. He argues that small states develop an early and specific type of democracy and economy (not that this is the only way a country can be democratic)1.
From de Tocqueville, we have associations ➟ social trust ➟ liberal democracy. Writing his Memoir on Pauperism in 1835, however, he could easily be disagreeing specifically with Fox’s writing: “they associate not to enjoy life but in order to find the means of living.” The italics is my own: it emphasises the ‘pragmatic’ aspect of de Tocqueville’s explanation.
It leaves an obvious question: why do we get associations? Why were the English early to adopt birdwatching clubs and democracy?
For Katzenstein, small/isolated society ➟ economic incentives ➟ (early types of) democracy. The hypothesis is an interesting one. Like Fox, he thinks that being small or isolated matters, but like de Tocqueville, argues that this creates important economic incentives, not social ones. Really, Katzenstein’s argument is more about economics than democracy. To flesh this third hypothesis out, we should recourse to Why Nations Fail, by legendary economist Daron Acemoğlu and political scientist James Robinson. They argue that societies in less hospitable landscapes are forced to develop inclusive political and economic institutions, like liberal democracy, earliest.
So we have three arguments.
First, that isolated countries desire the structured socialising of associations. I put this argument in red.
Second, that economic factors cause associations, which cause liberal democracy. This line is in blue.
Third, that small countries prosper with democracy due to economic reasons but not associations. This one’s in green.
To explore these hypotheses, I pulled data from the V-Dem ‘Varieties of Democracy’ dataset.
CSO, or Civil Society Organisations, participation index, which measures how involved civil society organisations (read: associations) are in the public sphere. Things like policymaking, how many people are involved in them, and how inclusive they are. It covers the years 1789 to 2020.
Participation, a pure measure of participation in associations, again covering 1789 to 2020.
Non-political organisations: associations, but excluding groups like trade unions and political parties, and focusing on things like book clubs, bird watchers, or charities. It covers 1900 to 2020 and is calculated differently, so I use this variable separately.
Higher values mean more associations. More details in the footnote2.
Fox — social dis-ease
It’s important to stress that the idea connecting isolated societies to social dis-ease is not her main argument. Much like de Tocqueville, the main contribution of Watching the English is simply the observation, not the explanation. But if this hypothesis is correct, we should expect to see higher values for associations in the smaller and/or more isolated countries. I first looked at ten countries that are rich and democratic today, and looked at their average CSO and Participation scores across three time periods, each running to 2020, from 1789, from 1949, and from 1993, respectively.
In the historical case, the more associational countries are all small/isolated and Western. By contrast, France and Germany are both large and continental, and score poorly. It’s important to remember that at this time the USA was not the modern behemoth, but more a young experiment and a wild west. Switzerland, mountainous and isolated from it’s sea-level neighbours, consistently scores highly.
But this only holds for European (or ‘Western’) countries: Japan and Taiwan are both islands, but score very low. Japan was a monarchy historically, and Taiwan was subject to Chinese and Japanese empires for much of its history, but Ireland was also an empire’s subject too: to Britain, until 1922, whilst New Zealand also lacked full autonomy. Not being democratic doesn’t explain why Asian island countries lack associations, particularly because this continues in the 20th and 21st century charts, as both countries become wealthier, and as continental Europe catches up to Anglo-American associationalism. (See this footnote for a historical chart of the same data from 1789 to 1913.3) Taiwan sees a similar trend, but Japan does not.
We can see the European factor quite clearly in the below chart, covering a range of countries. It’s ranked in order of their association Participation score, averaged from 1789 to 2020 (top), and second from 1950 to 2020 (bottom). All the small countries are highlighted in yellow, and it’s the European ones that completely dominate the top. Bigger European states like France and Spain are middling, mixed in with small non-European states, whilst larger, non-European states are at the bottom.
Historically, it looks as if being European and small is key. Contrary to Fox, it seems that size or isolation alone is not enough for a society to spontaneously create a vibrant network of associations. Perhaps what matter is geographic size, not isolation or population; Japan was as populated as France in the 19th century (and subsequently larger), and the data for Indonesia, not shown here, is also low.
We might be tempted to substitute the European criteria for wealth. After all, you don’t go bird watching if you’ve got kids to feed. But I’m not convinced. Associations popped up just as frequently in highly agrarian Ireland and isolated Iceland as they did in wealthy Britain. As the Irish historian Tom Garvin notes, Irish associations like the agrarian (read: poor) Ribbonmen and the culturally-sophisticated Gaelic League formed everywhere and vigorously (and often in response to imperial British rule).
Overall, small or historically isolated countries like Japan and Britain may have a social ‘dis-ease’, but Japan shows that this does not necessarily lead to lots of clubs and structured socialising. Only small European states really developed associations.
de Tocqueville — associations: correlation or causation?
The emerging hypothesis is that associations began as a feature of small European countries, such as Switzerland and the UK. Britain’s population was almost a third of France’s and half of not-yet-united-Germany’s in 1800! All of Scandinavia and Ireland score highly, each small European countries, whereas continental Europe does not.
Next, the hypothesis goes, Britain exported an association mindset internationally through conquest and colonisation (whereas Spain, France, and Portugal did not), and more speculatively, this ultimately spread across continental Europe and the more capitalist parts of Asia as Anglo-American ideas subsequently permeated the world and countries became more (neo)liberal.
The below chart lends weight to that hypothesis. It’s more of the above, using the Non-political organisations data. I looked at what countries scored in 1912, before WW1; in 1950, after WW2; in 1993, after the fall of the USSR; and in 2020, the most recent figure. High scores reflect high participation in things like book clubs and charities.
I added a global spread of nations to the ten used above. They cover South America (Chile, Uruguay), Asia (Australia, China, Hong Kong, Japan, South Korea, and Taiwan), Africa (Morocco, Tunisia), and the usual suspects in North America and Europe. Few of them are poor, and there’s a spread of autocracies and democracies.
I realised after making the graphic that I should have included Nigeria, Africa’s largest country and economy. It scores 0.87 in 1912 and 1950, tripling to 2.4 after independence by 1993, and rising to 2.9 today, making it one of the countries with the most clubs and associations across all periods. As a former British colony, it adds support to the Anglo-American hypothesis. But no way am I typing all those captions again….
Reading the captions is not required. There are three main insights:
British Empire countries consistently score highly: Australia, Ireland, New Zealand, the UK, and the US; Hong Kong is an exception. Marked by the 🇬🇧 flag.
Scandinavian countries always score highly. Marked by the yellow and red Kalmar flag.
Countries that were colonies of non-British empires score poorly, at first: Chile, Morocco, South Korea, and Taiwan (also Uruguay, sort of) — even though they’re all pretty small. They improve at varying rates.
One other correlation is very clear: membership of non-political organisations correlates significantly with whether your country is democratic, authoritarian, or subject to someone else.
South Korea, Chile, and even Morocco (still a monarchy) have all improved on the associations score since freeing themselves from colonists and dictators.
China is a good example of this, highlighted in the burgundy box on the left. In 1912, the Qing dynasty had recently collapsed, and two thousand years of Imperial rule was replaced by a young Republic of China thanks to figures like Sun Yat-sen. China wasn’t democratic, but it aspired to be, and Sun’s theories sought to blend those ideas with Legalist (read: state-led) traditions that had been practised in the country for millennia. Civil society flourished; the government had little to fear from people talking to their fellow citizens.
By 1950 however, ongoing discontent and civil war proved too much. Mao Zedong took over the country a year earlier in 1949, and, seeking to suppress dissent, Chinese citizen-led associations fell off a cliff. China’s aspirational republicans fled to Taiwan, where they started again. They struggled, and the island was a dictatorship for decades, but peacefully democratised by 1993. Its civil society has been on an upward trajectory ever since.
After becoming democratic, new democracies saw their associations grow to the same level as old democracies, whose scores are pretty stable and high, Meanwhile, non-democracies have consistently low scores.
How do we interpret this correlation between democracy and associations?
What is critical for de Tocqueville’s argument is that associations cause democracy.4 A common argument is that associations build trust and a shared character, which in turn helps build a democracy. This is part of de Tocqueville’s argument.
Egypt has always had a vibrant civil society and under General Hosni Mubarak had a genuine tolerance for a liberated press. This played a key reason in why the 2011 Arab Spring began in Egypt, rather than anywhere else.
Lisa Anderson was President of the American University in Cairo in 2011 and is a specialist in Middle Eastern politics. She writes that protests against the police violence that began the revolution “would have been unimaginable in Tunisia”, which lacked the “culture of deep communal bonds and trust” that was present in Egypt.
However, it was Tunisia which emerged as the most successful revolution. She adds that Tunisia “had the Arab world’s best education system, largest middle class, and strongest organised labour moved".” Civil society helped start the protests, but it didn’t finish the job5. Egypt now just has a new general in charge.
And consider this map about trust from Our World in Data. As Anderson writes, trust is a big part of this argument. Associations are meant to build trust, so citizens trust each other to build towards democracy. But democracies seem no more likely to be trust their citizens than other countries. Vietnam, Kazakhstan, and Belarus have very high trust, whilst France, Spain, and Brazil all rank low on this scale.
In summary, I don’t think associations are that important for liberal democracy.6 They are able to undermine autocracies, less so promote democracy. Meanwhile, there is little evidence that associations contribute towards democracy.
Katzenstein — small states
So where are we left?
Let’s remind ourselves of the map:
Small/isolated countries may have a social ‘dis-ease’, but Japan shows that this does not necessarily lead to associations. Only small European states really developed clubs and societies. Fox is out.
Even if a social ‘dis-ease’ did lead to associations, associations don’t seem to lead to democracy. It might make autocracies less stable, but they don’t actively help build democracy. I’m not convinced by the de Tocqueville claim either.
Yet Fox is partially correct in that isolated states do form more associations. It’s just that they also tend to be small and European.
It’s interesting that these small European states became democratic first. With the exception of highly democratic Switzerland, they’re all Nordic or English-speaking. Fox notes that in anthropology, these are all bucketed as ‘temperance cultures’.7
The takeaway is that whilst it’s unlikely that small/isolated countries are particularly associational, trusting, or democratic, Katzenstein’s link from ‘Small States’ to liberal democracy could be true.
For him, the small states of interest are in postwar Europe, from Austria and West Germany to Norway and Sweden, and contrasts them to “large industrial economies”, which include Britain, the US, and Japan.
I generally find this very convincing. But I’m also going to save it for next month, because I want to tie it into the future: what it means for technological change and the economies in which it happens.
The general idea is that small states cannot compete with bigger exporters internationally. As a result, they pursue niche export markets. But to do this, you need to develop a highly skilled workforce because niche export markets are found in innovating upon old technology rather than creating new tech (both these assumptions of where niche export markets are and how to pursue them are both contested; I half agree with them). Taking them as given, the next step is then that you need political institutions which give more rights to the working class so that they can influence policymaking in a way that makes high skills and training possible, as well as an environment where businesses built around these high skills can thrive, rather than descending into a race to the bottom. I strongly agree with this. Giving more rights to the working class means more democracy and liberty, whilst developing higher skills leads to (more equitable) prosperity. This whole framework is generally known as the Varieties of Capitalism framework. See Hall & Soskice (2001) for the definitive overview.
In all cases, the actual values are irrelevant; they are standardised relative to each other and to the mean. My Civil Society Organisations variable covers the years 1789 to 2020, and is known as v2x-cspart on page 52 of the V-Dem codebook PDF. The participation variable is known as v2csprtcpt on page 195. For non-political organisations, see v2canonpol on page 230.
It’s worth noting that this data definitely isn’t perfect; I also cut out Iceland, Ireland, and Taiwan because there wasn’t enough of it.
On the other, we might argue that democracy leads to associations, not the other way round. We saw how democracy unlocked the gates to associations in China, until dictatorship put a stop it. If forming clubs is the ‘natural state of the world’, then it makes sense that they only disappear when the government wants them to. We saw how associations thrived in Ireland and New Zealand, even under imperial British rule, which fits with the notion that associations exist by default, except when dictators actively suppress them.
But this struggles to explain why there are so many democratic countries, like Japan, with so few associations. To argue that clubs form by default is a strong claim, and there are so many democracies and other countries where associations are uncommon, which makes that claim unconvincing. And if you rule that out, but you still want to argue that democracy leads to associations, you’d have to argue that democracy actively causes participations in those associations, which seems unlikely: the whole point of democracy is that people are not forced to do things like that.
It can even be the case that associations and a strong civil society works against democracy, rather than for it. There’s a great paper about how Weimar Germany had a flourishing civil society, but this helped Hitler’s rise, rather than supporting the nation’s burgeoning democracy, because the German dictator brought associations under Nazi control. Think Hitler Youth.
We might argue that democracy and associations are only correlated. Instead, democracy and associations might both be caused by something else, like our pragmatic economic incentives. We know that cooperatives formed in Northern Europe in response to social and economic strife, and formed in North America around the same time as the development of democracy.
Temperance cultures are defined by their relative and historic abstinence from alcohol relative to other Western states. This might just be a descendent of Protestant work ethic ideas, which aren’t that useful (you’re just left with ‘so why did some countries become Protestant?’). But I’m keen to bring in more lessons from anthropology to political science. Anyway, the Irish drink less than the stereotype: