King’s College London recently hosted a lecture given by their Professor of Public Policy and Political Economy, Mark Pennington. In that lecture, he expressed a methodology which is both empirically and normatively sound for constructing and limiting political institutions. The talk was about the ‘project of political economy’ and more specifically, ‘realistic idealism’ which
... rejects a narrow focus on questions of efficiency because it recognises that policy prescriptions that take no account of moral constraints are not worthy of serious consideration. Equally however, it recognises that those forms of ethical or moral theorising that ignore real world practicalities of economics and politics are not worthy of consideration either... It draws on the work of economists, political scientists and philosophers [...and] recognises the importance of certain constraints that reflect the human condition.
Much of the debate on freedom of speech has been done over and over and I my main aim is not to go over recent controversies (although I will touch on them in the final part). Whether ‘offensive’ words should be censored is Very Boring Discussion #1. Whether Charlie Hebdo is racist is Very Boring Discussion #2. I want, rather, to go back to the basics and apply the above methodology to free speech: what does the empirical literature tell us about political/economic realities and the human condition in relation to the rationale for free speech?
This is an area which I am not as comfortable talking about as terrorism and international affairs so please do refer me any literature not cited here. I am myself am not entirely persuaded by some of the arguments I’m making, but I offer them to be refuted – or strengthened. The length of this post is explained by two factors: first, I do not want to leave claims – no matter how obvious – unsupported because it’s simply a bad methodology. A lot of what follows, unfortunately, many of you will already agree with.
Second, I am using this post as an opportunity to express my wider political and ideological views. I have tried to push as much of the post into endnotes where they don’t affect the main argument made here or are relatively undisputed. However, my primary interest is the talking about studies so cutting down hasn’t gotten me that far. I’d like to thank the Hated Sam Bowman, Ben and Pseudoerasmus for giving me comments on this post prior to publication. They have (differing) disagreements with this post but were open to reading it.
This post will likely be in three parts – most of which have now been written. The bibliography will be in the final part as I’ve been putting it together as I go along alphabetically and can’t be bothered to subdivide. I will make the entire thing available as a PDF at the end as well.
By way of summary, here is the argument I’m putting forward:
- The argument that a market place of ideas works to help individuals find the truth is not empirically robust.
- The argument that a market place of ideas works to help progress in human, political and social development over long periods of time is valid.
- The argument that certain kinds of speech are harmful is overstated by some but by no means entirely invalid.
- There are normative positions we should adhere to that have a radical impact on free speech and public policy.
- The argument from infallibility, coupled with the incompetence of the state, is the strongest argument in favour of free speech.
- The language that is used by both free speech advocates and detractors is misguided.
1. Dead Dogmas: Because Screw You, That’s Why
In line with his utilitarian background, Mill’s justifications for free speech were, at the very core, consequentialist:
If the opinion is right, they are deprived of the opportunity of exchanging error for truth: if wrong, they lose, what is almost as great a benefit, the clearer perception and livelier impression of truth, produced by its collision with error.
I am sure everyone is aware of this idea: the market place of ideas produces truth and progress. There are a number of ways this claim can be assessed – and we can assess them by talking about Mill’s more specific rationales. The first is that truths are held as ‘dead dogmas’ i.e., ‘as a prejudice, a belief independent of, and proof against, argument.’ The way to challenge these beliefs is to allow for the market place of ideas to operate. It’s worth noting that Mill is not saying that this will work on everyone:
I do not pretend that the most unlimited use of the freedom of enunciating all possible opinions would put an end to the evils of religious or philosophical sectarianism... I acknowledge that the tendency of all opinions to become sectarian is not cured by the freest discussion, but is often heightened and exacerbated thereby... But it is not on the impassioned partisan, it is on the calmer and more disinterested bystander, that this collision of opinions works its salutary effect. Not the violent conflict between parts of the truth, but the quiet suppression of half of it, is the formidable evil: there is always hope when people are forced to listen to both sides...
The most famous study in this area is by Lord et al (1979). They presented individuals who disagreed/agreed with the death penalty with evidence that contradicted their view. (Amusingly, the subjects of the study were offered a choice of ten cards that supposedly had a different view/study but the cards all said the same thing). The graph below shows results for the subjects who received the pro-deterrence study first and the anti-deterrence card second.
In the first scenarios, both opponents and proponents became more likely to support the death penalty. But once further details were given, the opponents of the death penalty became more entrenched in their oppositionist position. In the second scenario, despite being given anti-deterrent card, opponents became even more entrenched in their positions. The question as to which side of the debate is right isn’t really the issue here. Rather, its the idea that free speech may not allow people to hear both sides so that they can find the truth. Interestingly, this study is not a unique finding. A more recent example comes from et al (2015). This study used more issues than Lord et al and divided the subjects into ‘liberals’ and ‘conservatives.’ They found:
...conservatives had significantly greater motivated resistance to persuasion than liberals in the conservative-dissonant condition, whereas liberals reported significantly greater resistance than conservatives in the liberal-dissonant condition. Interestingly, in each condition, both conservatives and liberals resisted dissonant science messages more than their counterparts in the ideologically neutral condition regardless of their ideology
Beyond controls, institutional trust [in the scientific community] was lower in both the conservative dissonant (climate change/evolution) condition, and the liberal-dissonant (fracking/ nuclear power) condition as compared to the ideologically neutral condition. In other words, there is a net drop in trust of the scientific communication around politically contentious science in both contexts compared with the more neutral science topics...
Biases, then, may actually work to undermine the logic of the ‘dead dogma’ rationale. However, a number of caveats have to be added to this line of research. First, the ecological validity of these studies remains a very high concern: these people are being tested in a situation where their preferences have already stated may not really match how they generally act in the world. Second, Mill acknowledged that the argument was for ‘disinterested’ parties. It is an empirical question to the extent there is an undecided segment in relation to each particular issue.
Third, I have argued in the context of Iraq and WMD that you can be reasonable despite being wrong. This is not a relativistic argument but one about the nature of reasoning and the complexity of the world around us. When assessing these studies and indeed, the subjects in them, we must always be conscious of what this means for political economy and free speech. In particular, pluralism need not descend into relativism. This might sound abstract but I will return to this point below. Fourth, and linked to the first two points: an argument can still be made for progress generally over longer periods of time even if it cannot be made for individuals.
2. Progress: Civilisation and the Spirit of the Age
Of course for Mill, free speech was not just important in challenging the dead dogmas of individuals. When Mill was nineteen years old, he had adapted his father’s ideas about free speech in an essay, ‘Law of Libel and Liberty of the Press’ in the Westminster Review. His father, James Mill had written that ‘when all opinions, true and false, are equally declared, the assent of the greater number, when their interests are not opposed to them, may always be expected to be given to the true.’ Mill in his Westminster Review had changed the argument somewhat: ‘truth, if it has fair play, always ends in triumph over error, and becomes the opinion of the world.’
There are two possible interpretations of this. (1974) says that James Mill ‘expected only the greater number would give asset to the truth, John Mill was confident that the opinion of the world [i.e., an even greater amount] would coincide with the truth’ (p.35). I think this interpretation is mistaken: Mill is not making a quantitative claim here. In On Liberty Mill states in the third chapter:
The spirit of improvement is not always a spirit of liberty, for it may aim at forcing improvements on an unwilling people; and the spirit of liberty, in so far as it resists such attempts, may ally itself locally and temporarily with the opponents of improvement; but the only unfailing and permanent source of improvement is liberty.
Mill’s argument is then that over time free expression can is the driving mechanism of progress. In the examples that follow below, I will be trying to give evidence about the link between said example of progress and free speech but there will also be a more general causal argument at the end of this section. If you think I’m ignoring genetics, institutional factors or material causal factors in these examples, please humour me for now.
2.1 Political Freedom
The increasing amount of liberal democracy in the world is a trend that has been fairly well documented. Mueller (2014) wrote an article in which he looked over the empirical record since The End of History had been written. As he notes:
[Note: chart taken from Pinker (2011) to illustrate the wider trends]
Overall, according to Freedom House measures, the percentage of countries that are electoral democracies rose from 41 when Fukuyama’s article was published in 1989 to 61 in 2012, while the percentage deemed fully free (a high bar) rose less impressively from 37 to 46.
An even more recent study somewhat mitigates this rise of democracy by noting that in the last few years there hasn’t been as much improvement recently. and Way (2015) look at the several measures of democracy and find:
According to leading democracy indices such as Freedom House and Polity, then, the world is more democratic today than it was in 2000 (and considerably more democratic than it was in 1990 or any year prior to that). Even if we take the mid- 2000s—often cited as the beginning of the democratic recession—as our starting point, three of the four indices show either no change or a slight improvement.12 Only Freedom House shows a decline between 2005 and 2013, and that decline (from .63 to .62) is extremely modest.
To move on to the causal link between the progress we’ve made in terms of liberal democracy and free speech, Mueller is probably right when he says that
Democracy’s rise has, it seems, essentially been the result of a 200‐year competition of ideas, rather than the necessary or incidental consequence of grander changes in social, cultural, economic, or historic patterns. It has triumphed because the idea that democracy is a superior form of government, ably executed and skilfully promoted—or marketed—at one point in the world’s history, has increasingly managed to catch on.
By way of an elaboration I want to briefly talk about two books. The first is Leon Aron’s Roads to a Temple: Truth, Memory, Ideas and Ideals in the Making of the Russian Revolution, 1987-1991. I have reviewed this book elsewhere but the core idea is that freedom of speech allowed ‘every institution – political economic and social – to be subjected to trial by truth and conscience’ (p.51). It is following this process of self-discovery and criticism that surveys showed ‘solid majorities favour some key features of liberal capitalism’ (p.32-3). It then when people accept ‘alternatives to the current view’ does a ‘pre-revolutionary situation... become a revolutionary crises’ (p.20).
The second is Bernard Ideological Origins of the American Revolution which focuses on the pamphlets that were ‘the literature of the Revolution’ and ‘convey scorn, anger, and indignation’ that were ‘, speculations, theories’ which ‘were not mere mental gymnastics’ but were they provided the ‘grounds of resistance.’ (p.8, p.18, p.231). shows that the American Revolution was
... above all an ideological, constitutional, political struggle... [and] intellectual developments in the decade before Independence led to radical idealisation and conceptualisation of the previous century and half of American experience, and that it was this intimate relationship between Revolutionary thought [as expressed in the pamphlets] and the circumstances of life in the eighteenth-century that endowed the Revolution with its peculiar force... (-xi).
These two books are not quirky artefacts of the historical record. Chenoweth and Stephan (2008) look at ‘data on 323 violent and nonviolent resistance campaigns from 1900 to 2006.’ They find that ‘in the face of regime crackdowns, nonviolent campaigns are more than six times likelier to achieve full success than violent campaigns that also faced regime repression.’ Their data show that no non-violent campaign (i.e., an exercise of free speech) that garnered the participation of 3.5% of the population failed to reach its goal.
Free speech, then, is vital to denying authority legitimacy, and then finally removing the regime. We can also see this by looking at how curtailing free speech is vital to maintaining autocracy. One of the most interesting studies I’ve read recently is and (2015). They posit the following model:
The new-style dictators can brutally crush separatist rebellions and deploy paramilitaries against unarmed protesters. But compared to previous regimes, they use violence sparingly. They prefer the ankle bracelet to the Gulag. Maintaining power, for them, is less a matter of terrorizing victims than of manipulating beliefs about the world.
The way that beliefs are manipulated is through bribing informed elites (which could send out messages to the populace about the incompetence of the dictator) or controlling the outflow of information either by censoring existing media or by putting out his own propaganda. The study notes that to do these things requires increased taxation which thereby reduces the wealth of individuals and thereby exacerbated his incompetence score. By manipulating information (i.e., curbing free speech), the citizenry cannot conclude that the dictator is incompetent and therefore cannot overthrow him. ‘The most fundamental failure of totalitarianism was [and is] its failure to control thought’ (Fukuyama (1992), p.29).
(2012) is another fascinating study which tries to explain the lack of democracy in the Middle East by looking at government regulation of religion (GRI). This is defined as ‘the restrictions placed on the practice, profession, or selection of religion by the official laws, policies, or administrative actions of the state’ (hence the relevance to this post about free speech). Does the level of GRI explain the democratic deficit in the Middle East? First, they found there clearly was a difference in levels of GRI: ‘the average GRI for non-Muslim countries in 2005 is 2.36 out of 10, it is 6.42 for Muslim-majority countries.’ Second, they ran multivariate analyses – the key finding of which is:
Model 2 adds GRI to the previous model, showing that when government regulation of religion is considered among the determinants of democracy, the variable for Muslim majority countries is no longer statistically significant... In Model 1, the coefficient for Islam is −5.1. Once GRI is added in model 2, the coefficient is reduced to −2.39 and loses its significance.
You can explain the democracy gap between the Middle East and the West by accounting for the extent to which a state has free speech curtailed. The point that is making is that laws which curtail free speech are used against political dissenters. Its why, for example, Muslims are more likely to be targeted rather than non-Muslims in laws against proselytizing. For reasons that I have gone into elsewhere, Islamic institutions are often the only vestige of civil society left in the Middle East. This leads to a fear of ‘competition and threats to [autocrat’s] power, [and then] repressive governments restrict the religious arena in order to forestall the rise of political opposition.’
I should note that the internet should not be confused with free speech. Rod and (2015) note that authoritarian regimes are more likely to introduce internet access to their populations (see table above): an increase of ten points in Press censorship predicts an increase of up to 0.33% in Internet penetration. In addition, internet access does not have any link with democratisation:
within the low penetration group there are 11 movements towards democracy in nine countries and zero movements towards autocracy... while there are no movements towards autocracy in the low penetration set of countries, there are eight episodes of autocratic regime change in six countries in the high penetration group... there are almost twice as many countries (nine vs. five) experiencing movements towards democracy in the low penetration group as in the high penetration group.
The reason behind this is because the internet is suppressed and used as a tool of repression. The authors of the study used Saudi Arabia as a case study to explain their data. I’m sure most people know the kind of controls that the Saudi have: you need a permit to post a YouTube video, there are block pages (which, even if you wanted to use a proxy to get around, citizens are still deterred). The idea of seeing the internet as part of liberation technology is one that depends entirely on the institutional framework.
There has been a decline in violence over long periods of time – not just in terms of tribal cultures vs. modern cultures, but also in recent history. This is a point that I think is rather well settled. There is declining inter-state conflict and declining intra-state conflict, not just in terms of numbers of conflicts but in terms of battlefield deaths:
Pinker summarises the huge declines and progress we have made in a recent article in The Guardian:
Research institutes in Oslo and Uppsala compiled datasets of global battle deaths since 1946, and their plots showed an unmistakable downward trend. The per-capita death rate fell more than tenfold between the peak of the second world war and the Korean war, and then plunged an additional hundredfold by the mid-2000s.... Other datasets show steep declines in genocides and other mass killings. The declines are precipitous enough that they don’t depend on precise body counts: the estimates could be off by 25%, 100%, or 250% and the decline would still be there.
Am I saying that these historic declines are all because of free speech? Absolutely not. The trend seems to be so well established that those who seek to undermine the trend look to non-human causes of the decline. The most popular of which is the improvement of medical science – ands its actually not totally baseless. (2014) looks at this argument in more detail in relation to battle deaths. is arguing that there hasn’t really been a decline in violence, simply that military medicine means fewer people are added to the body count. The way that he measures this is through wounded-to-kill ratios (i.e., the amount of the attacked actually dying). An increase in the ratio means that fewer people are dying even if the same number are attacked – and that’s exactly what he finds:
But can’t actually explain all of the decline. In an ideal world, he would use the data above and ‘regress some function of time as well as the wounded-to-killed ratio on battle deaths’ – but the data is so sparse he can’t do that. What he does do though is to look at the decline in fatalities vs. the decline in casualties. If the decline in battle deaths is being driven by medical improvements, we would expect to see no decline in casualties, and a decrease in fatalities. Except we don’t:
Clearly, the improvements in modern medicine can explain some of the decline. But it can’t explain it all. I can do no better than ‘FAQ’:
First, before the late 19th and early 20th century, most medicine was quackery, and doctors killed as many patients as they saved, yet many of the declines I document occurred before that time. Second, many forms of violent crime move up and down in tandem—for example, rapes and robberies went up in the 1960s and down in the 1990s, just like homicides—so it’s unlikely that any of these trends simply consist in a constant amount of violence which has been reallocated from deaths to injuries thanks to quick-acting EMTs. Third, while medical technologies have improved, so have weapon technologies.
Fourth, advances in medicine can only move the numbers around for the statistical sliver consisting of the victims of violence who are injured so severely that they would have died with even with the primitive medical care in the past, but not so severely that would have died even with the advanced medical care of the present. Yet many of the declines are from scorched-earth campaigns of violence in which no amount of medical care could have reduced the death tolls to current levels—Mongol invasions, deliberate sieges of cities (in which doctors, even if they were around, would not have been allowed in), over-the-top frontal assaults into machine-gun fire, Dresden, Hiroshima, carpet-bombings, the deliberate killing or starvation of prisoners of war.
(See below on health for why this is also linked to free speech). And obviously, aside from medical technology improvements, there are many other causes. It is beyond the scope of this post (and my current understanding) to even attempt to explain the historic decline. But I think free speech does explain some of the decline. I think this for two reasons. First, in terms of direct mechanisms free speech allows the spread of norms which eschew violence – especially for more violent countries. and (2014) much like everyone else above notes a decline in battle deaths and conflict intensity but their data focuses on Africa:
On the whole, however, Africa is less war-torn than at any time in the past, which runs contrary to widespread perceptions that exist even among foreign policy experts.... the evidence suggests that despite neo-Malthusians fears, by most measures life on the continent is improving. War is becoming less of a threat to the life of the average African than emerging threats like traffic accidents or diabetes. Nor have realist fears of predatory wars and wholesale remaking of the map of Africa come to pass.
Whats more interesting, however, is one of the reasons they give for the decline. The mechanism they describe relies on the free flow of information:
Every modern state is part of an interconnected international society, where ideas and norms spread with unprecedented rapidity... Twenty-first-century Africa exists in a complex, globalizing society whose members have been slowly abandoning the recourse to warfare. Its leaders and its people would not be unaffected by such powerful global trends.
[These trends, in the form of] conflict resolution norms in the global north affect decisions in the south. Success breeds imitation; the behaviour of prestigious states will be copied... It would be hard for Africa to remain immune from a fundamental transformation in beliefs regarding warfare in broader international society, particularly with modern communications reducing isolation. It is difficult for leaders to credibly claim war is a useful, necessary option when the notion is rejected elsewhere.
Moving onto the indirect reasons: as argued above, free speech helps degrade the legitimacy and remove non-democratic regimes. Non-democratic regimes are simply more violent to their own populations, to other populations and more prone to civil war. Before I get on to the empirical literature for these claims, take a look at this graph from Pinker (2011) on the rates of deaths in genocides and mass killings (p.407):
The first thing to note is the unsurprising fact that these regimes are not liberal democracies. In fact most of the spikes emanate from non-democratic, totalitarian communists like Mao or Stalin. Imagine what the death toll would have been if these regimes had been removed by the power of free speech. I'm not interested in the histories of particular dynasties, I’m interested in trends and interestingly there is a consistent finding:
For thirty years, quantitative research has supported this relationship. Repeatedly, democratic political systems have been found to decrease political bans, censorship, torture, disappearances, and mass killing, doing so in a linear fashion across diverse measurements, methodologies, time periods, countries, and contexts From this work, one could conclude that with every step toward democracy, the likelihood of state-related civil peace is enhanced. (Davenport and Armstrong (2004))
Davenport (2012) notes that ‘all investigations over the last 40 years [show] democracy has decreased repression’ but there is one ‘controversy that exists within this work concerns exactly when along the spectrum of democratic governance this influence ‘kicks in’.’ See here for the three models. I would note that the most well supported are the linear model (more democracy, less repression) and the threshold model (reach a threshold of democracy, less repression). Davenport and Armstrong (2004) undertake an analysis of 147 countries from 1976 to 1996 and they find:
Across databases and methodological approaches, our statistical investigation leads us to conclude that there is a threshold of domestic democratic peace. Below certain values, the level of democracy has no impact on human rights violations, but after a threshold has been passed (varying in accordance to which measure one is considering), democracy decreases state repression.
The mechanism I’m suggesting is free speech --> democracy --> less violence. The same trend exists in terms of inter-state violence. The literature on the mechanisms for this trend is somewhat divided. The usual mechanism expressed in the literature is the idea that the electorate is not prone to support wars and democratically elected leaders act in accordance with that desire lest they be removed from office. Far more interesting, however, is the mechanism put forward by and Weeks (2013). They suggest that ‘shared democracy pacifies the public primarily by changing perceptions of threat and morality, not by raising expectations of costs or failure.’
[The perception of threat felt by democracies toward other democracies is lower because] democracies expect other democracies to externalize peaceful norms in the same way and therefore trust that they will not be attacked by other democracies... [The moral argument is based on] people in democracies... feel morally reluctant to overturn policies that the citizens of other democracies have chosen freely
This is probably the most obvious way that free speech helps. The scientific method is based on the idea of peer review helping us obtain the Truth, a Truth that has a real world impact (see above). Advances in medical science require the principle of free speech to operate. In The Great Escape: Health, Wealth and the Origins on Inequality, Deaton (2013) charts the improvements we’ve had in health outcomes across the world:
... in the fifteen year period from 1950-55 to 1965-70, the “less developed regions” of the world saw an increase in life expectancy of more than ten years, from 42 to 53 years. By 2005-10, this increased by another thirteen years, to 66 years... [For sub-Saharan Africa,] the gap between it an Northern Europe has narrowed, from 31.9 years in the 1950s to 26.5 years (p.107-8).
There have, of course, been declines: HIV/AIDs and Mao’s famine are two notable declines in the graph above. The graph actually understates the impact of Mao’s policies as a number of accounts show that life expectancy fell from 50 in 1958 to below 30 in 1960 (p.39). (Incidentally, I would really like to write a review of The Unknown Cultural Revolution which tries, hilariously, to sanitise Mao’s record but nobody cares). What is the cause of these health gains? One part of it is, as mentioned, the free flow of ideas. As Deaton notes:
The knowledge that cigarette smoking kills has saved millions of lives in the past fifty years... That germs cause disease was new knowledge around 1900, an professionals and educated people were the first to put that knowledge into practice... New ideas, new inventions, and new ways of doing things are the key to progress... Scientific advance – of which germ theory is such a singular example – is one of the key forces leading to improvements in human well being (p.7, 9, 100).
It has helped in other ways as well. There has been a historic decline in homicide around the world:
[Graph from Pinker (2011), p.75]
...the data confirm the notion, now hardly controversial among historians of crime, that homicide rates have declines in Europe over several centuries. Typical estimates referring to the later Middle Ages range between 20 and 40 homicides per 100,000, while respective data for the mid twentieth century are between 0.5 and 1 per 100,000... the evidence is so consistent, the secular decline so regular, and the difference in levels so large, that it seems difficult to refute the conclusion of a real and notable decline ( (2001)).
There was of course a rise between 1950 an 1990 but that has since been in free fall across the West. Its also worth bearing in mind that, as Eisner (2012) notes the ‘very low rates of homicide found across most of Europe during the late 1950s, when the period of sustained increase begins, should probably be seen as a rather exceptional phenomenon.’ Since 1990 there has been a significant decline. Much of this decline, in terms of overall trends, is attributable to medical advancements. Harris et al (2002) suggest that if the medical technology in the 1990s was equivalent to what it was in the 1960s, the homicide rate could have been up to three times higher (see also (2014) above). The improvements in technology require innovation which necessarily rely on free speech.
2.4 Economic Freedom and Income Growth
More than any other area I am talking about in this post, the literature on economic freedom is the least known. In particular, the trends of economic freedom are not spoken about – particularly by those who prize it most. Economic freedom is the idea that taxes, government interference, labour market regulations should be low and measures of free trade should be high. It is, in essence, an important element of a market economy. (2014) is the only study I know that looks at global trends of economic freedom. And he finds that 2007 is the peak of economic freedom:
The academic literature is fairly clear that economic freedom is good for economic growth. De et al (2006) after reviewing the literature (see Table 5) state that ‘it is clear from these studies that EF seems to have a positive association with growth’ and ‘those studies that deal with the problems of model specification and sensitivity in a more rigorous way also find that there is a positive growth effect from EF.’ My favourite study in this area is et al (2013). They look at the relationship between IQ and GDP per capita and how it is moderated by economic freedom:
That is, the relationship between GDP per capita and IQ is significantly affected by the level of economic freedom a country has. Economic freedom is not just good for growth, it is also good for labour market shares (Young and Lawson (2014)), poverty reduction (Hasan et al (2003)), lower unemployment and labour market participation (Heller and Stephenson (2013)), income (Ashby and Sobel (2008)) and on and on the list goes. (2015) looks at how economically free countries fare against economically unfree countries in recessions. Given the studies above, his results are unsurprising:
Across these crisis episodes, it is evident that crises tend to be substantially deeper in countries with relatively little economic freedom than in countries with high freedom: the peak-to-trough ratio [i.e., the percent drop of real GDP per capita from its pre-crisis level to the last year of the crisis] in the former is 10.3 percent while it is only 3.4 percent in the latter.
... An increase in economic freedom of ten points, or slightly less than a standard deviation, is associated with a decline in the peak-to-trough ratio of four percentage points, or half a standard deviation. This is associated with a reduced recovery time of approximately ten months.
The only caveat to attach to these results is that there is some literature which holds some attributes of economic freedom (low regulation, low government interference, high level of protection property rights, low taxation) are more or less important than others. In addition, there is a lot of publication bias ( and (2006)) but the effect seems pretty well supported ( (2007), and (2011), Compton et al (2011)).
The rise of economic freedom to the highest point it has been since 1850 should not been seen apart from the decline of poverty and the rise of income. Economic growth, fostered by economic freedom and innovation (see below) has led to a phenomenal amount of improvement in income – even if we just look at the period from the 1960s to now, from Deaton (2013):
Almost all of the darker circles [i.e., those that represent 2010] are above and to the right of the lighter circles [i.e., those that represent 1960]; since 1960, nearly all countries have become richer and their residents longer lived. This is perhaps the most important fact about wellbeing in the world since World War II: that things are getting better, that both the health and income parts of wellbeing have improved over time.
Aside from impacts on health, this translates into a pretty good trend of poverty reduction. My favourite study in this area – which shows both the link between economic growth/poverty reduction and the radical poverty reduction itself is and Sala--Martin (2014):
Africa is reducing poverty, and doing it much faster than we thought. The growth from the period 1992–2011, far from benefiting only the elites, has been sufficiently widely spread that African inequality, if anything, declined during this period... the driving force that appears to explain the substantial reduction in poverty between 1992 and 2011 is economic growth.
In the UK, the bottom quintile’s income rose 93% between 1977 and 2011 (ONS (2013) quoted in (2015), p.67-8). The UK median income is the same now as that of richest 10% in the 1960s. The bottom 10% average income is the same now as median in the 1960s. In the U.S., post-tax incomes of the bottom quintile have risen 48% in the period between 1979 and 2011 (CBO (2014)). This understates the huge progress we’ve made in economic well being:
You, oh average participant in the British economy, go through at least sixteen times [1,500 percent] more food and clothing and housing and education than an ancestors of yours did two or three centuries ago. [The material ease the average American has gained is] 1,700 percent, a factor of nearly eighteen [in the time between President Monroe and President Clinton] (Bourgeois Dignity, McCloskey (2010), p.48-49).
However, as Piketty notes in Capital we should be cautious of using averages:
...the only way to accurately gauge the spectacular increase in standards of living since the Industrial Revolution is to look at income levels in today’s currency and compare these to price levels for the various goods and services available in different periods... When family budgets and lifestyles change so radically and purchasing power varies so much from one good to another, it makes little sense to take averages... (Kindle Locations 1611-1612, 1646-1647)
He notes that for foodstuffs like milk, butter and eggs, there has been a increase in purchasing power whereas with meats there has only been a fourfold increase. ‘French purchasing power expressed in terms of oranges increased tenfold, and expressed in terms of bananas, twentyfold’ (Kindle Location 1630). Nonetheless, it still is indicative of how much we understate the improvement in well-being – especially because ‘the things poor people want appear to be dropping in price faster than the stuff rich people want.’
The link between free speech and economic growth can be made with a general argument that the ideas of free markets, economic freedom has won out. But it can also be made in terms of the importance of innovation to growth. Innovation requires a culture of innovation which ‘new and sometimes radical ideas were respected and encouraged, heterodoxy and contestability were valued, and novelty tested, compared, and diffused if found to be superior by some criteria to what was there before’ (Mokyr, (2012) quoted in et al (2015)).
Rather than provide one of the boring studies that shows this proposition to be true, I want to look at et al (2015). et al look at the links between religion, politics and growth. They find that ‘religiosity is significantly and negatively associated with innovation per capita [as measured by patents]’ (see, in particular, figures 2a-b 3a-b). They then propose the following model to explain these result:
If allowed to diffuse widely [innovation] will produce, at the start of the second , advances in practical knowledge and technology that raise TFP [total factor productivity] from at to at+1 = (1+[Equation])at [i.e., TFP significantly increases] ... [Also] new scientific findings that contradict the professed doctrine and sacred texts’ statements about the natural world... tend to shake and weaken the faith of religious agents.
Not all innovations and knowledge will lead to ‘belief eroding’ (BR), some are ‘belief neutral’ (BN). In the following passage, ‘β₁bt+1Gt+1’ refers to the utility that religious agents derive from organised religion:
If allowed to disseminate, a BR discovery will reduce the utility β₁bt+1Gt+1 of religious agents, through both its direct erosion of their faith and the ensuing reduction in Gt+1 [i.e., belief in religious public good institutions like churches, spiritual assistance]: If this loss more than offsets the gains to be reaped from higher TFP, the government, representing here the religious majority, may want to block, censor, deny, restrict access to, etc. the new knowledge [but will have to set up an apparatus to do such things].
The first period (on the left hand side referred to as ‘t (even)’) shows the options that religious agents have open to them. They have to choose to block or not block BR ideas (note that taxes will have to be levied prior to any block to establish the apparatus to stop new ideas). If a decision is taken to block, TFP at ‘at+1’ is ‘at’. By contrast, TFP rises where a decision is taken to not block. The point is not that religion is necessarily bad for growth (that relies on a number of other factors) but that religiosity can explain a state’s policy toward knowledge which can, in turn, explain economic growth.
Importantly, the environments for innovation that we create in the civilised states have an impact in uncivilised states. After the Cultural Revolution wiped out scientific and technological output in China, it took a great many years for China to return to producing a high number of heavily-cited academic papers and a high number of patents. As Freeman and Huang (2015) note that the ‘global mobility of people and ideas allowed China to reach the scientific and technological frontier much faster than if it had gone down a more parochial path.’ The period in which China made a real great leap forward in science and technology was the period in which Chinese places at Western institutions rose dramatically too:
And that is arguably not a coincidence. I consider the link between innovation and economic growth to be settled and will therefore not dwell on the subject (see, for example Gordon (2014), Piketty (2014) at Kindle Locations 1634-1635, and this wonderful essay from Anton). In summary, two big determinants of economic growth (economic freedom and innovation) have in the case of the former spread because of free speech or, in the case of the latter, are inherently based on free speech.
On Liberty was published posthumously and so remains the final word of Mill. Many fringe left historians who provide ahistorical accounts of Enlightenment thinkers ignore the changes that Mill underwent in order to present him some kind of authoritarian maniac. The gist of one of these early essays, Civilisation, is that ‘power passes more and more from individuals, and small knots of individuals, to masses: that the importance of the masses becomes constantly greater, that of individuals less.’ This isn’t much different from the tone of On Liberty until he uses this as an example:
This is a reading age; and precisely because it is so reading an age, any book which is the result of profound meditation is, perhaps, less likely to be duly and profitably read than at a former period. The world reads too much and too quickly to read well. When books were few, to get through one was a work of time and labour: what was written with thought was read with thought, and with a desire to extract from it as much of the materials of knowledge as possible. But when almost every person who can spell, can and will write, what is to be done?
Mill goes on to say that there are only two occupations in which the individual has power, and again, ‘regrets’ something that is an anathema to On Liberty:
[There are] but two modes left in which an individual mind can hope to produce much direct effect upon the minds and destinies of his countrymen generally; as a member of parliament, or an editor of a London newspaper. In both these capacities much may still be done by an individual, because... the number of participants in it does not admit of much increase. One of these monopolies will be opened to competition when the newspaper stamp is taken off... and the influence of any one writer in helping to form that opinion necessarily diminished. This we might regret... (my emphasis)
It’s strange to read Mill saying that the increase in ideas will lead to bad outcomes. What is even more illiberal was Mill’s ‘remedy’ for this is purported issue. There could be no remedy ‘while the public have no guidance beyond booksellers’ advertisements’ and so he proposed the following ‘resource’:
...the resource must in time be, some organized co-operation among the leading intellects of the age, whereby works of first-rate merit, of whatever class, and of whatever tendency in point of opinion, might come forth with the stamp on them, from the first, of the approval of those whose names would carry authority.
By the time of On Liberty, he dropped such illiberal ideas; and the evidence above shows that he was right to. We have seen an insurmountable amount of literature rise, and with it, we have not lost our ability to determine what is useful. We have made huge gains. Above, I have tried to give specific mechanism for the progress that we have seen. There is also a more general mechanism. Matt Ridley in his Ted talk notes that ‘ideas have sex’ and its this ‘interchange of ideas, the meeting and mating of ideas between them, that is causing technological progress, incrementally, bit by bit.’ Zack Beauchamp in a wonderful review of The Great Escape (quoted above) for ThinkProgress makes a similar point:
...the new rationalism taught that “happiness could be pursued by using reason to challenge accepted ways of doing things, including obedience to the crown and the church, and by finding ways of improving one’s life, including both material possessions and health.”
[This] suggests that there’s a third type of innovation, beyond those of science and business, that propels humanity forward: moral advancement. Enlightenment politics were, after all, a product of Enlightenment morality. [For example, the] unanswerable moral challenge to monarchic privilege — “Who are you to rule us? Why are we not your equals?” — inspired the democratic, rights-respecting political systems
This moral advancement is relevant to many areas where we have progressed that I have not mentioned above. The decline of racism, sexism and homophobia are all things which free speech has been involved in. These are not merely generational changes but the result of an active engagement of ideas. The rationality of man necessarily means that we move in the direction of progress:
This argument is not without weaknesses. MacDonald argues in The Invisible Hand of Peace (which, I must admit, I have not read) that the extent to which a state engages in capitalist policies is a better predictor of peacefulness. If the book is correct, it shows just one of the mesh of connections between the factors I’m talking about. The decline of violence is link to both improvements in technology (discussed above and below), democracy (above). Economic growth is impacted by democratic institutions and health and vice versa. Much to Sam and Ben’s chagrin, democracy leads to high levels of economic freedom:
. (1) to (3) [show] the effect of democratization on [economic] liberalization is positive and highly significant at the 1% level, while making an unstable transition has a significantly negative effect at the 5% level... a shift to democracy is on average associated with a 0.26 higher change in EFW [measure of economic freedom] after 5 years, a 0.30 higher change after 10 years, and a 0.23 higher change after 15 years, relative to stable autocracies (Rode and (2012)).
Note that the dependent variable in Table 3 is ΔEFW (changes in economic freedom). Economic growth, economic freedom, democracy, health, technology and innovation - these factors are all linked. This mesh of things probably weakens the argument I’m making somewhat: how much of the progress can we really ascribe to free speech when () there are so many other factors and (ii) the progress-factors I list are interlinked causal factors. That’s a fair point to which I don’t have a complete answer. I think, though, it doesn’t eliminate the role of free speech. I hope to have shown the role that the free flow of information and ideas has for significant development and progress. But simply because free speech can lead to good outcomes does not mean it is not also a cause of bad outcomes, which brings us onto the next part: the harm that free speech leads to.
 Fukuyama (1992) provides a more general history which shows the roles of ideas:
‘The critical weakness that eventually toppled these strong states was in the last analysis a failure of legitimacy—that is, a crisis on the level of ideas. Legitimacy is not justice or right in an absolute sense; it is a relative concept that exists in people's subjective perceptions. All regimes capable of effective action must be based on some principle of legitimacy. A tyrant can rule his children, old men, or perhaps his wife by force, if he is physically stronger than they are, but he is not likely to be able to rule more than two or three people in this fashion and certainly not a nation of millions... It is clearly not the case that a regime needs to establish legitimate authority for the greater part of its population in order to survive. There are numerous contemporary examples of minority dictatorships that are actively hated by large parts of their populations, but have succeeded in staying in power for decades.... When we speak of a crisis of legitimacy in an authoritarian system, then, we speak of a crisis within those elites whose cohesion is essential for the regime to act effectively’ (p.15-16).
I think this emphasis on the role of elites to the detriment of force is misplaced but that doesn’t change his core argument that, in the spirit of both Aron and Bailyn, ‘there was a remarkable consistency in the democratic transitions in Southern Europe, Latin America... [With a couple of exceptions] there was not one single instance in which the old regime was forced from power through violent upheaval or revolution... [rather] it was ultimately made possible by a growing belief that democracy was the only legitimate source of authority in the modern world’ (p.21).
 As The Economist noted: ‘Recent data from the Oxford Poverty and Human Development Initiative show that the link between the MPI headcount and income per person is just as robust as that between the extreme-poverty headcount and income per person.... There is, unsurprisingly, a strong and negative association between the extreme-poverty headcount and GDP per capita; as countries grow richer we expect the incomes of the poorest to increase.’
 Pinkovskiy and Sala-i-Martin (2014) is an interesting study because, it seems to me, that it poses problems from all leading schools of thought regarding economic growth. In particular, note that
All classes of countries, including those with disadvantageous geography and history, experience reductions in poverty. In particular, poverty fell for both landlocked as well as coastal countries; for mineral-rich as well as mineralpoor countries; for countries with favorable or with unfavorable agriculture; for countries regardless of colonial origin; and for countries with below- or above-median slave exports per capita during the African slave trade
This paragraph poses problems for Jared Diamond;s (geography), Acemoglu and Robinson’s(institutional) and typical human capital accounts of economic paths.
 Bénabou et al have a nice brief section on innovation cultures in the Muslim world. As they note, the initial period of Islamic rule did not have any major aversion to innovation and turned up a whole host of new knowledge. It then started to wane, the printing press being one example:
...printing –especially in Arabic and Turkish– was strongly opposed throughout the early-modern and modern periods. In 1515, Sultan Selim I issued a decree under which the practice of printing would be punishable by death. Printing only started in the Islamic World at the beginning of the 19th century, partly due to the need for defensive modernization against the West.