GovAI Webinar #2: Carles Boix and Sir Tim Besley on Democratic Capitalism at the Crossroads: Technological Change and the Future of Politics
Introduced by Allan Dafoe
Featuring Carles Boix
and Sir Tim Besley
Okay, welcome everyone. I hope you can all hear and see us. Today we have the privilege of having Carles Boix and Sir Tim Besley talk to us about insights from Carles’s recent book, Democratic Capitalism at the Crossroads:Technological Change and the Future of Politics. Carles is the Robert Garrett Professor of Politics and Public Affairs in the Department of Politics and the Woodrow Wilson School of Public and International Affairs at Princeton University. He has published important work in political economy and comparative politics, particularly around the role of institutions in shaping economic growth and inequality. Carles’s work has been especially impactful on my career choice, I’m happy to share in 2003 when I was considering what fields to move into — economics, political science, sociology, — I read Carles’s book on Democracy and Redistribution and I was at the time and remain deeply impressed by the sweep and importance of its argument, but also its theoretical parsimony and breadth of empirical support, I highly recommend it. This book, in fact led me to see political science as a discipline compatible with asking big important questions such as, as will be discussed today the impact of advances in AI on political institutions inequality.
Following Carles’s talk, Sir Tim Besley will offer some reflections on the theme. Tim is the School Professor of Economics of Political Science and W. Arthur Lewis Professor of Development Economics in the Department of Economics at LSE. He is also a member of the National Infrastructure Commission and was President of the Econometric Society in 2018. Tim has published in top journals in economics, and I’m also happy to see in top journals in political science on topics spanning political violence, state capacity, economic development and their interactions.
Following this, we will have a 30 minute discussion. At the bottom of the screen you can see ask a question, please do click on that and articulate your questions and comments. And then also vote up those questions that you most think worth discussing. We may not get to too many or any of them, but I will certainly try to look at them and incorporate them into the conversation. Okay, so with that out of the way, it’s, again, our pleasure to have Carles introduce some of the themes and insights from his book.
Okay, so, thank you so much for your inviting me to this and for this very kind presentation and thank you to Tim for the you know, being willing to discuss the book, and of course, to all those attending. I think It’s going to be a bit difficult, like talking to a dark room. And normally we are used to seeing what the audience thinks, at least when we give seminars or talks. So I’ll try my best. I hope that the future is not like this and it’s going to be like a short shock to all of us and we can reconvene in a normal way. So I’m going to use some slides and to let me look for them.
So basically, this is a book that I published last year. And it’s basically driven by a double motivation. The first one is what I think is a long term intellectual question, which is the compatibility of democracy and capitalism. This is a long running question that in the 19th century was basically answered in a very pessimistic way, both by the right and the left. So Marx thought democracy and capitalism were incompatible. Having one person one vote, could not be together with the protection of what he called the interests of the bourgeoisie. But it was also answered in a pessimistic way by conservatives and in fact, by liberals like john Stuart Mill, who made the point that they would only be compatible, provided the population was educated at a high level. Then after these periods of pessimism, what we see is in the 20 century, at least In the second half of the 20th century, a period of optimism, the coining of the term democratic capitalism, the possibility of having representative democracy, markets or regulated markets, but it’s still free markets and a kind of strong welfare state. And then today, what we see is a moment of questioning of the compatibility of these two things. And so that’s where the book comes. And these links, the sort of long term intellectual question with today’s politics, that’s the second motivation, which I perhaps I think it’s what has driven most of you to come to to attend this talk. And the politics of today, at least in the advanced world, is one where you we see a lot of mistrust towards politicians, growing abstension, polarisation and the rise of what many call populist parties.
So this process or this ideational change that I’m talking about during the 19th century, 20th century and today also coincides or comes parallel to economic change. So what I’m showing here basically is over time for a few countries for which we have some data the US, Britain and Japan, the evolution of their level of income inequality measured by the Gini index. And what we see is high levels of inequality in the 19th century, a sharp decline by the middle or the end of the first third of the 20th century, and then approximately around 1970, 1980 a growing inequality in those countries, so three periods. And my answer in the book is that in a way, this is driven by technological change, which then has consequences on the labour market and in politics, these technological changes basically, in let’s say industrial capitalism, produced or generated by a stretch for more productivity, an increase in efficiency leads to basically a process of capital labour substitution that results in what I would call different production models, especially in what regards to what is the kind of labour that is complimentary to capital. So, in the book I distinguish, and I discuss three moments that are not precise but that kind of dominate each century, what I call the Manchester model, where basically what we see is a process of mechanisation. At least of some industries. The paradigm of them would be the textile industry with the use of mostly unskilled labour, the decline of demand for artisans, an increasing use of unskilled labour, that then coincides or leads, there is discussion on that in the literature, with declining wages and a more unequal distribution of income.
By the beginning of the 20th century or at the end of the 19th century, a set of technological changes such as electricity, what is called the use of parts that are interchangeable that can be used in many different industrial processes, and the production system that has a sequential layout, literally, assembly line in Detroit, or in the batch production machine, for example, tobacco machines that basically result in the decline of very unskilled labour and semi skilled labour becoming complimentary. These transformations in the labour market, I claim in the book are associated with growing wages for everyone with a distribution of income this equalising and with the formation of a class that in the British sociological literature was labelled the affluent worker or having an affluent working class. Then by the 1970s and 80s a bunch of changes, mostly an increasing computational capacity and a decline in computing costs leads to a new transformation that I center in Silicon Valley.
So, everything in the book moves west from Manchester to Detroit, from Detroit to Silicon Valley. That all these changes in computation lead to an automation of basically routine jobs, and in a way also fosters a new globalisation a real globalisation that includes a set of countries that were in the periphery of capitalism into the world economy, China, East Asian countries with important effects on wages in Europe and North America. The complimentary labour at this stage is high skilled labour. And as volunteers such as the board or have shown the hollowing out of the labour market, a polarisation of the labour market, a decline in what were considered middle class jobs. Here are good jobs in the past. So a growing inequality.
These changes have an impact on politics. In the 19th century, the incompatibility of democracy and capitalism led to basically very restrictive suffrage regimes. But in the 20th century, what we see is after a period of strife and World War One and Two and the crisis of the 30s, a process of democratic consolidation, what sociologists and political scientists called the end of ideology period, and base and pivotal politics with parties competing for the median voter. This has been in a way replaced today by conflictual politics, by polarisation and by what some call a possible real crisis of democracy and democratic backsliding. I’m going to focus on the rest of the talk on basically this political side of the story. Which basically occupies the last two chapters of the book.
Let me say that when we look at democracy, at least in advanced countries, which are basically where I pay attention to in the book with some references to developing countries at the end of the book, and at the end of this talk, when we look at the politics of democratic countries, what we observe is a process of, as I said before, depolarization. So here I show how left or centre left, certain right, centre right parties, on average, were located in terms of their position in ideological left right, using party manifestos for which we have information about available at least since the end or the beginning of the Cold War period. What we see is a process as I said of depolarization, so the lines, both of them converge to words, more moderate positions, both the centre right parties here in blue, and the social democratic parties in red, they become very similar or relatively similar on average for of course, in the late 1980s. But what is also very interesting about this graph is that, after that process of convergence, they remain basically, at the stable and close to each other.
This is comes in contrast with a process of political disaffection at the mass level. So, here what I show is the proportion of people that respond that “politicians care about what people like the respondent, like me,” what we see is a growing or a decline of trust in politicians. The longest series is the US series represented in black, for which we have data going back to the mid 1960s. At that time, about 60% of Americans thought that politicians care about what people like them thought. There was a decline that coincided with the Vietnam War, a slight increase in the 1980s, and then a big our fall in are in the 1980s. With a very short spike related to the to the Twin Towers attacks at the beginning of the 2000s. Today, only 20% of Americans think that politicians care about what they think. But the same process happens in Germany in blue, in France, in red in Britain, already at low levels in the 1970s. Now at only 10% of British voters think that politicians care about them, or at least this was for 2014. It’s not the same everywhere. So here I show a case where the trend goes just opposite. This is the case of Finland, where trust is at about 40%. So much higher than any of the big countries. There are others small countries like the Netherlands that have have a similar levels of of trust, or at least in 2015.
This process of political disaffection is correlated or comes with growing abstension. So here I show in green, the levels of abstention in the US excluding the South which is a different or was a different thing. Abstention was extremely high as a result of exclusion of African Americans in Europe is in black. And what we see is that extension till the early 1970s was at 15%. extremely low. At that time, the study of abstention in the scholarly literature was a non issue. So it’s very hard to find any one that was interested in that topic. But today, abstention in Western Europe stands at about 33% not that different from the US in the East Coast and in the West.
The abstention is concentrated in particular sectors. So here I show three countries Finland, for which we have real data because they have a registration system that allows us to track all individuals and surveys from France and the United Kingdom and I show the rate of abstention divided by cohorts. So young, middle and senior. And then by income quintile, just here, the top quintile, the middle quintile, and the bottom quintile. And what we see, basically, is a extremely high level of abstention among young people, but also among those that are in the low income levels. So certainly groups that are have experienced, I think, economic shocks, and that in some economies the young are more excluded from labour markets somehow are turning out much less.
When we think about the stability of party platforms I was showing before and this kind of growing alienation of the population the question basically brings to us is a puzzle, why is that, given the degree of dissatisfaction that was building up among voters for a long time, European mainstream parties, so centre right centre left parties, those in favour of the consensus of democratic capitalism didn’t react. In part they did, but they did not as much. And it’s surprising because when we look at the support among the electorate, what we see is that starting in the 1970s, the percentage of support for centre right, and centre left parties, which was at about 70% had been like that, since the end of World War Two started to decline today. They together have the support of about 45% of the electorate, so it’s a big decline. It’s corresponds with increasing abstention, of course. But the question is why didn’t these parties respond?
And I think that there are several explanations to that. It could be that fiscal constraints such as a growing population of pensioners, globalisation prevented them constrained mainstream parties from acting more strongly to in response to this decline in trust, but I think that part of it was also a question of electoral incentives. When we look at the proportion of votes that mainstream parties got in elections, basically until the crisis of 2007, 2008, their level of support was very stable at around 80% and it has only been in the last 10,15 years that as a proportion of voters, they have started to bleed votes. This has led to a changed political landscape.
So, here in a sort of very stylized way I represent how we may want to think about electoral politics during what I would call the trade capitalism or the golden age of capitalism. Or if you will, the post war period. The graph has two dimensions one is about compensation. So basically taxes from low to high. And then there is a vertical dimension these would be globalism, trade, immigration, from being very much against globalism, so at zero to being very much in favour of globalisation. And then I represent the location of voters in blue, it would be were more or less the location of middle class voters. And in red the location of working class voters, this is these curves are not the circles or are not indifference curves is just more or less representing what is the bulk of the middle classes and the working classes at that time. And then two parties right and left located are closer to the let’s say, median voter. Here clearly the preferences of voters are only different or heterogeneous along the compensation dimension. Globalisation at that time was not that important. Again, I’m talking about advanced democracies. And it was not important because there was globalisation in the sense of high integration of advanced countries. But the developing countries were not really competing with industrial workers in advanced countries, this has changed progressively. And there has been a process whereby the middle classes have become more heterogeneous, with some part of the great work of those voters moving towards embracing even globalisation even more, those and some are less. And then the working class affected by all the transformations of automation and globalisation, moving towards a position that is protectionist, if you will. Here, what I’m talking about is preferences. So in a way these preferences have been framed, or structure the narratives have been constructed by politicians themselves.
And so in this new world, what we see is the growth of an alternative instead of the old right and left of the past, what we see is the growth of this thing called populism, which is a term that I refuse to use without quote marks, because it’s a difficult term to define. And I would rather talk about these as anti globalisation or nationalist movements. And so in response to this movement that insists on the second dimension, the trade and immigration and mentioned then the response, I think, at least from the left to get many of its voters back seems to be a more polarised position. So that would be a movement from L to L prime. So here we would have if we think about the US, Trump versus Sanders or something like that. So that’s how these economic transformations have transformed the politics of the advanced world. And when we look at some data on vote for populist parties, what we see is, again, I here I show the proportion of people voting for populist parties in the mid 2000s divided by income quintiles. So, basically, in the bottom quintiles, the proportion of voters voting for populist parties, more than double sometimes triples the vote for populist parties among those at the top income quintile. Of course, this is mediated by electoral institutions in proportional representation systems, populist parties have had a much easier life in terms of getting the vote and becoming a viable party that is voted by by the electorate in countries like the UK with a majoritarian system. Can parties like the UKIP didn’t do well in elections. But of course, everything exploded in the around the question of Brexit referendum.
So, let me now kind of finish by thinking a bit about or talking a bit about what’s next, which is kind of the discussion that is all over the place. And here what we find there’s a division between techno optimists and what we want to call techno pessimists. This is not a new discussion, we go back to the 19th century, the 20th century, and we have the two positions there. So Keynes, as well known from his well known as, say, economic possibilities for our grandchildren, has a optimistic position about the effects of technology. And basically, consider that we three hours a day, we probably will have enough and we will be able to do many other things, such as what Marx thought communism would bring. So, you know, hunting in the morning and fishing in the afternoon or the other way around. We also have the pessimistic Marx, at least for capitalism, who thought that at the end of the day, technological change would lead to monopolistic capital to the immiseration of the working class and two columns of workers going down into Manhattan and Silicon Valley, I suppose, to burn everything.
So what is my position? Well, really, I think we do not know the future. Right? So here I’m showing a graph that I find I don’t know if funny but at least entertaining and revealing. This is data from a paper by two authors Armstrong and Sotala that look at papers and books with predictions about when artificial intelligence may replace human activity. And so what I plot here is I use their plot, in fact, is when did they predict in the horizontal axis that AI would replace human activity in the sense of when did they publish the paper and then in the vertical axis, what was the year they predicted AI would replace human activity. And what we see is predictions all over the place. That go from for the latest from 2020. So papers published in 2010 thought that we would already be replaced. To some that put the date at the end of this century. In fact, there are some authors like the father of the singularity movement, Kurzweil, who have made predictions for different years over time. So if we do not know what may happen, the only thing I think we can do is to consider different scenarios, as they are defined by preparing parameters.
That’s what I do at the end of the book. I consider four things that I think will determine where are we going to go. Three of them are mostly economic, if you will, and one is political. Those parameters are labour demand, the supply of labour, how concentrated capital may be, and the responses, the political responses to all these in the north, but also in the south.
So I will just spend my few minutes last minutes on political responses by saying that my assumption through the discussion is that what has changed from the 19th century. So if you remember the first graph I showed was about the evolution of inequality from high to then declining, and now, again, growing. What differs today is that in the 19th century, even the most advanced countries, the industrial countries were poor in comparison with what we are now, right so, a country like the US had a per capita income in 1870 at around $2500 are in 1990 constant dollars Today, the per capita income is about $30,000. So this change, in a way has to have an important impact in terms of the room of manoeuvre that we have to do many things that we could not have done two centuries ago or 150 years ago.
So what may be the impact of these technological changes? As I said at the beginning, there has been a lot of talk about the crisis of democracy about democratic backsliding. And so when we look at the data, what we find is that in terms of the number of democratic breakdowns since 1800 till now, there were no democratic breakdowns basically in the 19th century, because there were no democracies. Then we see a lot of democratic breakdowns in the 1920s, 30s and then again during the 20th century, but when we look at us today, and given that there are many more democracies, what we see is at least till 2015, which is the last year for which I have reliable data, democratic breakdowns have not gone up. So that’s an optimistic thing or information or piece of information we have. What about the probability of democratic breakdowns? Here, what I show is what is the probability that any democracy today breaks down four different levels of income from these, the horizontal axis is in thousands of dollars, so 5000 10,000 15,000 20,000 and the probability of democratic breakdown here is calculated as the number of democratic breakdowns over democracies for different levels of income from all the data we have in the last 200 years.
And for poor countries, the probability of democratic breakdown at any given a year is about 6%. For very rich countries, it’s about basically zero percent. So, in principle, the question is that when we look at economic development, more developed countries are more likely to be democratic, or at least for those that are democratic, not to break down.
And so, there are many different explanations that have been given. It could be that economic development leads to a as a result of declining marginal utility of income makes the rich indifferent about the the distributional mechanisms that come with democratic voting, it could be that the economic development is correlated with a set of ideational commitments, such as toleration, and so on that then lead to a more willingness to have democracy or, and that’s the explanation that I think I prefer, which is that without excluding the others, but I think it’s more important is that development has led to a different incentive, a different structure of the economy has led to at least till now, more equality that then has been used on the conflict, social conflict, and has allowed democracy to flourish.
So basically, this argument in some way to be the quick arm runs like this. In the old times, before industrialization in most countries were governed by monarchy and nobility, what also would refer as stationary bandits. And as a result of the way they governed, the distribution of wealth was unequal. And in that context, democracy was impossible. Those elites would blocked the introduction of representational mechanisms. Then with industrialization, a new system appears what we may want to call open economies with free markets and with technologies that equalised growth. So 20th century capitalism. When we look at the distribution of cases in terms of [inaudible] country years for those cases, we have information with the vertical axis being the Gini index and the horizontal axis being income.
In the 20 century, what we see is a mass of countries, that among poor countries, there is a lot of heterogeneity in terms of levels of inequality. But for high levels of development, what we see is much lower levels on average of inequality. That is the story, I think of the 19th century and especially late 20th century capitalism. And so of course, the question is, what is I think in the back of our minds, is, what if these open economies, so, industrial capitalism, as we know it, free markets generate increasingly unequal economies? What are the chances of democracy in this context of yes, much higher levels of prosperity, but also much more inequality? Now, I think that it’s difficult to know it will happen, it will may well be that this inequality may lead to elite capture and to democratic backsliding and so to mass resentment, so it will basically undermine the foundations of democracy. On the other hand, and I here want to be optimistic or at least sound optimistic is that provided that the allocation of returns is thought to be fair, so markets function in a way that do not give advantage to those that have more. And given our already high income levels, it may result in democratic stability.
What do I mean by that? Well, we know that in lab settings, participants generally allocate resources in equal shares, but we also know that they distribute them unqually when in these lab settings when people are engaging tasks by that require effort. We also know from observational data from surveys, that inequality to people is unacceptable when it is at high levels, but some inequality is acceptable. So if we maintain a system that is fair, and then democracy, even if there is inequality may well survive. And I think that in the past democracy was impossible, meaning before the 19th century or before the 20th century, because besides most of the population being poor, the distribution of assets was extremely unfair. The second part of it is that because we have high incomes, this should allow us to develop institutions to respond to shocks to compensate losers to make sure that people are helped to adjust to these new technologies. Now, for this to happen for this democracy to be of quality and to not disappear, we need institutional reforms, in my opinion, to prevent an excessive concentration of economic and political power to avoid crony capitalism. Here, I think that some democratic accountability mechanisms, such as, of course, financial reform in the US, but also the introduction of ideas, such as the electoral voucher should help. I think that breaking up large companies, I don’t want to get into the Zuckerberg Musk dispute, but I think that making sure that those that have become very advantaged or benefited from these technological changes need to be regulated and perhaps, are divided into several companies that also should help.
A third thing that I think is not that discussed that electoral participation matters, that the fact that a lot of people are not turning out and that these abstainers basically come from low income young people, these are changes the median voter and changes the median representative and has an impact on policy. So changes in participation are fundamental mobilisation may be fundamental to reverse the things that are not good for democracy.
And finally, a fourth thing, which is country size. I think that small countries for some reason, because of the tight relationship between representatives and and voters are better positioned to respond to the excessive concentration of economic and political power of the capitalism of today. This is something that I just suggest. It’s not that I’m completely sure, but I think it’s something that should be investigated a bit more by all of us.
So let me finish by saying that in the book, what I challenge is the standard story. We have a standard story that says that modernization led to the end of history, and instead what I suggest is that we should think about capitalism as having different technologies, with different effects on employment on wages. And so these may explain the kinds of political honeymoons and political moments of conflict that we have seen in the last 200 years. So thank you so much.