Fri, 23 December 2016
What follows is an edited transcript of my conversation with Judy Stephenson.
Petersen: You're listening to Economics Detective Radio. My guest today is Judy Stephenson of Oxford University's Wadham college. Judy, welcome to Economics Detective Radio.
Stephenson: Thank you very much. It's nice to be here.
Petersen: So, our topic for today is economic history. Specifically we’ll be looking at some interesting research Judy has done on wage rates in the early modern period in London. This period is particularly interesting because it's the start of the Industrial Revolution which leads to a dramatic increase in the growth living standards and of technology and that trend of course is what has shaped our modern world and made it different from the world of the past. So, it's very important of course to understand this period if we want to understand the world as it is now. So Judy, start by giving us historical background. What was the world like in the period you study?
Stephenson: Well, I work mostly on researching London, so urban environments. And London is very developed in this period between about 1600 and 1800. And London becomes the biggest city in the world during this period and as the biggest city in the world it's hugely vibrant, some of the largest merchant houses in the world are there, banking is advanced and developing. Most of the occupations of London are tertiary or service sector, even at this early date.
The river is a huge source of both transportation and work, the port is where much of the capital, both physical and financial, from around the world comes through the city, and the professions and bureaucracy are well established in London in this period. It's growing at all levels of society, from the very poorest to the very richest exponentially. So, if you look at the population growth overall in the U.K. in the late 17th century from 1500-1600 to 1700, that actually is pretty much stable or slightly declining. But the population of London grows by a third or something in that period.
London is this hugely vibrant commercial social and cultural center and it's pretty much overtaken Amsterdam, which has come to the end of its golden age in the mid 17th century, right at this period. So, although the world more generally and in a wider sense can be typified by pre-industrial or agrarian values, London is very commercial in this period.
Petersen: Okay, so, if I were to get in a time machine and go back in time, maybe London would be more familiar to me, would seem, feel more modern than almost any other place.
Stephenson: I think it would be very familiar to you the way of getting around would be a sedan chair or a carriage. You can hire them on the street, in fact you send your boy out to get one. It looks very like Uber, it's a gig economy.
And most people working in unskilled, or who didn't have a trade or didn't have a profession or skill probably didn't have steady jobs. They thought of themselves as having work that they could rely on, but it wasn't wholly reliable and they definitely didn't have a contract that would keep them going, they probably didn't have many rights either. And they probably worked at two or three things and everything---the traditional literature about London in this period is one of inequality. So the very very poor literally scavenging on the streets among the smut because the streets were the sewers in those days, and the very very rich living in these incredibly grand environments with retinues and servants.
It's a golden age for the aristocracy after they had a pretty rubbish time in the 16th century. It's a golden age for the aristocracy, it's a golden age for art, for architecture, for all these things but it is also a period of desperate poverty and mortality. The plague doesn't die out in London until the end of the 17th century, but still very very high infant mortality and living standards are nothing like they become in the later 19th century, after they sorted out all those things. But from a commercial point of view, you might well recognise it.
Petersen: It's very interesting---and of course the whole period is interesting---but it's particularly interesting for what it becomes, really. The rest of the world starts becoming more like London, starting in this period.
Petersen: And so you study wage rate of some of the day labourers and the workers in that period. How have economic historians gone about measuring things and getting data that far back in the past?
Stephenson: Well, data on wages and prices for this period was originally gathered by a guy---Thorold Rogers---who was a 19th century historian who started collecting wages and prices in the mid 19th century and finished 40 years later, literally a broken man. These are seven volumes from around England and he basically went into any long run institution where there was an archive or records, as they were called in those days, and just noted the quantities and prices found in the books.
But it was a huge project way before the days of even print noting, before the days of an efficient typewriter, let alone a computer. It was pretty haphazard as to what he was actually recording but it's very accurate. But he tended to take down labour costs or wages as day rates, and what he mostly found were builders because he was in big Oxford colleges and places like Westminster Abbey which had buildings from the 13th century and had required a lot of building maintenance and surprisingly he didn't find many other wages.
So this way of recording had a sort of half dependence. These day rates because they were the only ones that people could find it was assumed that wages---wage rates are very hard to find but there's always good ones for builders---and it was assumed that builders were the same everywhere in terms of skill levels so these could be comparative.
And Arthur Bowley---who is known as the father of modern statistics, an economist and statistician again working in the end of the 19th century and in the early 20th century---used builders in his first attempts to think statistically about an average wage, an average worker, and to establish a real wage. And Bowley’s work is absolutely seminal in the history of statistics, econometrics, and economic history. And he used Rogers' and others' wage rates of builders. And this tradition carried on as other historians gathered more rates, like Elizabeth Gilboy in the 1930s, and then Phelps Brown and Hopkins used all these people's data when they came up with the seminal Seven Centuries of Building Wages in 1955.
And what Phelps Brown and Hopkins had done was they took all those day rates from the builders, and then they took a series of wages and prices and they created a basket of goods and they offset the wages against the prices and they came up with an index of the real wage or living standards across the ages. And this has been the standard for measuring welfare since 1955. And because it's very difficult to find wage rates for the 18th century for some of the reasons I spoke about a minute ago---not many people have jobs, etc etc etc---the dependence on builders' wages continued until, with the most amazing econometric and advanced econometrics techniques that Greg Clarke and Robert Allen were using, they still use that data from the 1930s.
I think the latest good index Jeremy Boulton made in the early ‘90's, where he collected about 2,700 observations of wage rates. The key thing to remember here is all of these wage rates came from bills in the archives of the institutions. So they’re not really wages. In fact they are not wages at all. So, I don't know if you've ever worked for somebody and been charged out by the day, have you?
Petersen: I have not, but my wife is charged out, she works in data science and yes, she gets one wage and she's charged out to other firms at a different rate.
Stephenson: And what she's charged out is higher, right? So, when I worked in advertising, I cost my clients about 1,800 pounds a day, I saw about 350 of that. What a bloody enormous margin, actually. You got to look at how IPG were not making a really stonking profit on that but you know there's overhead and those kinds of things.
Well, in the 18th century everything, but particularly in the building trades, that's exactly how you dealt with masons or bricklayers or carpenters or labourers. And any economy that has to organize production---and the building they were organizing was pretty big, the Great Fire of London destroyed the old city and was completely rebuilt in about a decade---there's some serious organizational coordination mechanism problems of making all that stuff happen. And the 18th century way of doing it is contract it out. Firms are a series of sub-contracts and so the way wage rates have been collected were the sums that were paid to contractors and what those contractors pay their men were substantially lower than those wages that Phelps Brown and Hopkins had used, or Robert Allen had used and Rogers and people have recorded.
Petersen: Okay. In your paper you mention Robert Allen and he had a hypothesis that based on these faulty rage weights that high wages in London were a contributing factor in kicking off mechanization in the Industrial Revolution. So, can you talk a little bit about that hypothesis and how your new look at the data has, I suppose, called it into question?
Stephenson: Yeah. So, Allen has made the most seminal contribution to the study of the Industrial Revolution. So, the Industrial Revolution is the savored big debate in economic history really and it's a favorite big debate for lots of parts or disciplines within economic history. The history of technology people like it because of the gadgets, the history of macroeconomics and supply and demand people like it because of the factor prices, the history of the organizational people and sociological people like it because of the institutions in the factories.
So it has this broad appeal for everybody who's interested in the economics of the long run. Essentially, the core issue around the Industrial Revolution is it's unexplained. Why did it occur in England before anywhere else? It's this naughty problem that had never really been adequately explained until the early 2000s. Then there were two competing---well not two competing but two complementary---explanations by sort of giants of economic history in the same period.
So, Bob Allen explained it through England being a high-wage economy and Joel Mokyr explained it through a series of innovations and enlightenment and how that brings about sort of an intellectual enlightment in scientific innovation. Allen’s theory was the economists’ theory and still is. And essentially what he proposed is that the high wages of England incentivized the owners of production to substitute capital for labour.
Essentially because of the way series are constructed when you take all those comparative wage series of Amsterdam, London, Milan, Florence, Madrid, Antwerp, Strasbourg, when you sort of put them all together as a real wage series in the long run, the English wages looked substantially higher by comparison, particularly after 1650. It looked like the cost of labour for capital in England was much higher than it was in the rest of North Western Europe or Italy, where you had the traditional textile industries and banking, where there was some quite advanced commerce in places. Allen argued that the high wage economy first of all created those incentives but that also it had created higher human capital and skills, attracted capital to it, to prepare England for industrialization in the long run. But that the trigger was induced innovation through relative factor prices.
And part of his theory also was that coal was cheap and available in England, which is very hard to argue that it wasn't, the coal in China is in Mongolia, the Dutch don't have any they've got coal in the Ruhr, of course. But you know coal has been at the center of English energy requirements for a very long time as Tony Wrigley has written about in a very distinct way actually in a lovely book called Energy and the English Industrial Revolution, which is the kind of thing your children could read.
So the relative factor prices between energy and capital and labour were unique in England is Allen’s argument. So, obviously if you find out that the wages are 20% to 30% to even 40% lower than Allen thought, that presents a problem for that theory.
Petersen: I believe I heard once that Germany had coal but it had to be transported over land and so was as good as useless to them before the age of the steam engine and trucking. Coal is really important. And so Robert Allen felt that high wages in London and in England were important but it seems like this issue of measuring the contract rate instead of the wage rate casts doubt on that, or even---does it close the whole gap between London and the rest of Europe?
Stephenson: Good question. And that really depends on what sort of organizational form or coordination mechanism was in place in other countries.
So,I've looked into this with Amsterdam and Antwerp quite a bit already. I've done some work with Heidi Deneweth who works on the Low Countries on economy and building particularly. She's at Ghent. And we're finding in the way that building is organized in Amsterdam, in London, is that in London very much the state has completely outsourced everything. So, the city doesn't employ people directly, that's too much hassle. It seems like the cost of management to something is very high in England because they outsource everything: the navy, the supply, the whole thing. Bits of the navy are integrated into it, but a lot of it, particularly the supply to it, is outsourced and all building is outsourced. Whereas in Amsterdam the city still employs people who are digging dikes, and looking after canals, and doing maintenance work on public buildings. Whereas in London the comparable projects which would be stopping London Bridge from falling down, or wharfing the fleet ditch and making these canals and things. Those are given to large contractors and the contractors are solely responsible for labour.
Whereas there is some relationship between labour and the city, people are directly employed in Amsterdam, this is indicative only and we need to do a lot more work on comparing contracts in the same types of organizations. And then there's a guy called Luca Maccarelli, who is an established Italian historian of the building industry and industry in Milan generally and he has looked at some of the data for the wages for Florence and Milan particularly and he has shown that the day rate was only part of the wage there. In fact the contractors were throwing food, bonuses, cash savings, access to places to stay, and all sorts of perks at workers to try and induce them to work. So the wage in Italy was probably a little bit higher. In fact, Mark Reilly has said that we've understated Italy’s by 15-20% and then the person who's done the most work on France so far is Vincent Geloso, who's shown that the Strasbourg wages are probably problematic.
But all this comparative stuff is at a really early stage. And we need people to get out into the field, the way I've been in the field in London, and look at more the form of employment and the form of the wage in those places. And really understand, the figures that we've got are they real or have they got other sort of recording factors like I've shown in London? So it's too soon to say although we started work on that.
Petersen: So, for the modern era we have people collecting data and they're making a big effort to collect the same data across time and across place. Surveys asking the same survey question to everyone, or government data and making sure it's collected in the same way every year but when we're going back to the past, of course there was no one in the year 1700 collecting data on Italy, and London, and Amsterdam, and all these different places. And so we have to stitch it together from what is available and often that's very different datasets.
Stephenson: Exactly, and different types of records. So, it may be the case that all the records are a bit skewed and you know there'll be a new schema once we have all the new data together that does reproduce the Allen’s story. And remember that we need to take the prices of goods into account. It's a real wage calculation he's done not just a nominal wage calculation. But until we've done that, what we do know is the living standards in England were not what Allen thought at the moment but you've got to do the whole comparative thing to know.
Petersen: So, how do you distinguish the skilled from the unskilled? How do you make sure you're comparing the same kind of labour?
Stephenson: That's a good question. Traditionally pretty much everywhere in Europe we've gathered two types of wage: a skilled wage for what we call craftsmen and craftsman are people who have completed an apprenticeship, who are qualified, that's the idea. So, a mason who has studied seven years in England---doesn't seem to be as long anywhere else---or a carpenter who has studied in the long run. So, who has invested time in the development of the human capital and acquired skills and then we think about the unskilled person as a counterpoint as being the labourer.
And this is another important distinction because you know building labourers are actually of two kinds: there's the completely unskilled guy. Actually there are three kinds: there's the completely unskilled guy who's basically just handing them nails or wheeling a barrel around. But then there's the more skilled or semi-skilled assistant who actually is doing a lot more than that, who is preparing the work for the craftsman, who knows which tools go with which materials and who is fully assisting a craftsman and they couldn't really do the work without them. And you call that semi-skilled. And then there's a labourer who is hired really for their brawn. They've got a premium for being extremely strong and what you tend to see in building accounts is people who are actually hired by the load. They get 2 shillings and 8 to move a ton over a day or something---and probably need more than one man to do that---but so there's a brawn premium in these labourers or unskilled.
And actually from Phelps Brown and Hopkins onwards we've taken this semi-skilled or brawn wage to be the unskilled wage, but these people aren't unskilled. Whereas the unskilled, the guy wheeling the barrel, or just picking out nails was paid a lot less than those. So, if the rate for the semi-skilled guy was 18 pence a day in 1700, the rate for the unskilled guy was 12 to 14. So you can see there's a considerable premium in here. That's another thing that colours our understanding of welfare because usually it's the unskilled or subsistence wage that the macroeconomist is interested in. They relate unskilled and subsistence even though they maybe should not. It's that unskilled wage that is an indication of supply and demand in the labour market, and the draw of that. So taking building labour to a semi-skilled to be unskilled leads to some problems because it implies that unskilled people in London could afford four times the subsistence basket of welfare goods in 1700, when actually they could barely afford two.
So, if you're going to use a welfare basket these rates have a real issue and the distinction between skilled is…
Petersen: So, the reason maybe we care more about unskilled wages is because that's the wage that you'd expect to see in other places in the economy. For instance unskilled work in agriculture or working in a shop or things that we don't have data for we can sort of guess because presumably there's a labour market and people have mobility and if there was too big a gap between wages for different unskilled jobs then people would move, they’d arbitrage away that difference. So your paper, it has some sort of case studies. You have data from particular construction projects. I thought those might be interesting to go through. So, one of them is the reconstruction of St Paul's Cathedral after the Great Fire of London, which is a massive project, could you talk a little bit about that?
Stephenson: Well, yes it's a famous project because the old St. Paul’s had stood since I think the 14th century. It was this you know cultural and emotional symbol for Londoners apparently, and it had been redesigned---the front had been redesigned---by Indigo Jones, the kind of father of classical architecture in England. And it was completely destroyed by the fire and this was a sort of symbolic task to rebuild and so Christopher Wren hailed the King, came up with the design and you know Wren is pretty much the father of modern architecture and he's this enormous intellectual as well as architectural figure, he's very much part of the enlightenment.
So the project lasted about 35-40 years, so they declared it finished in 1711 and the Great Fire was 1666 and it's still there today, absolutely intact, it survived the Second World War. So it's this incredible and very emotive building. The interesting thing from a work point of view is it's very much a craftsman's building, it's not an artist's building. So there is sculpture there, there is painting but nothing like a European cathedral like St. Peter's, St. Paul’s is very much a display of English craftsmanship and baroque style and most of it is stone faced.
So, I have these wonderful papers, which are the day books of one of the Master Masons, one of the contracting masons who built the south west tower on the west front. His name was William Camster, his father was also a contracting mason on a separate contract and in the network of masons who served, ran and worked. We’d ran over 30 or 40 years and he was on site for about 10 years of the project from 1700 to 1709 or so and some after and I have his day books right, years of this, where he records every single man that was working for him and what they paid him. So, it's got an appeal because you can go and see what they did---which is very rare---working on the 18th century that you get some wage records and you can actually see the product as well. So, it's quite nice from that point of view.
So, from an economist's point of view the interesting thing is the way that they contracted the construction because they just started out one contract at a time and then if it worked, they’d go "Yes. We'll do that again." So, they had these repeated idiosyncratic contingent claims contracting going on and on and on and obviously disputes arise and they resolve them, or people drop out and they get new contractors. But the whole thing is basically on a rolling contingent claims contract what Oliver Hart and Holmström said could never happen. Oliver Williamson would have had his head in his hands.
But the other notable thing is that the contractors financed this really because the Crown didn't pay them. It did pay them but the Crown and the city, they leveraged the coal tax but mostly people waited two or three years on contracts to be paid. So, the cost of financing that was just swallowed up by the contractors, it was in the price. And that's one of the reasons why you see a margin on labour and materials. But the interest costs for St. Paul's were as a total of the entire bill over 35 years about 20%, and very little of that had been lent by citizens and the city, a lot of that had come from the contractors themselves through just rolling over bills.
Petersen: That's interesting. So, we know not only what they were paying their day labours, but also implicitly we know the interest rate for that time.
Stephenson: We do. Yes, 6% for to and from the cathedral. Six percent on an annualized basis. Stephen Quinn and Temin and Voth have found higher rates, above 8% for some private lending around the same time. And it is likely that these contractors will have had to have done some private borrowing or lending within their networks to keep rolling this finance over. Because they will have bought the stone, they will have paid the carter, they will have paid the labours who are working for the carter, they will have paid the craftsman, so they may have well have to borrow to do all those things but 6% is what they got from the cathedral.
But the real question is then, so these networks of supply chains are surviving on that kind of finance. So really big contracts essentially on a very high level of trust or a very high level of interest. We need to do more work to find out which, but it does seem like these networks---because they repeatedly contract---they have good information and it's more effective than you would imagine those types of contracts to be.
Petersen: And of course they're contracting---it's the government paying for it ultimately right?
Stephenson: Yes, and it's financed through the coal tax which is also interesting. Bearing in mind the price of coal is relevant to development at this time. The coal tax was levied at a shilling a cauldron after the Great Fire to rebuild the churches for the city and then it was maintained through and into the Georgian period by parliament who kept sort of either adding to it or continuing it and apparently it was detested and greatly avoided.
But we definitely need some more research on how this work, and how people avoided it, and and what it did to coal consumption. Because you find in the accounts that the coal tax, they're expecting this much per year from it and consistently about 10 to 15% less comes in. So they have to turn to the city or to commissioners and people who might have money to borrow from them and tide it over. So financing the thing was unconventional.
Petersen: So, we usually think of government debt as being highly safe at least in the modern period but back then it may not have been.
Stephenson: Yes, and I don't know what the connection to other Treasury things are and Bank of England and everything. At the time it looks like it's just private between St. Paul's and the commissioners for St. Paul’s and either citizens or contractors and that it wasn't actually securitized as a state promise, but there may have been connections. It's something I haven't delved into enough.
Petersen: So, another construction project, in this case it's a maintenance project, is the famous London Bridge which of course in the nursery rhyme "London Bridge is falling down" which apparently was true. Can you tell me a little bit about that?
Stephenson: So, well London Bridge was it was built the end of 13th century and it's 19 stone piers across the Thames. It must have been the most fascinating and amazing structure, it stood for pretty much 500 years, but by the end of the 16th century in the early 17th century it is falling down.
And the Thames because this sort of development further up river as well, the Thames is actually a very strongly flowing tidal river at this stage and the force of the water force through those 19 piers is wearing away. So they built wooden starlings, so they built a wooden constructions they look like boats around the piers, trying to guide the water through and these of course made the problem worse and they made the waters faster. So to pass under the bridge in a boat at high tide apparently you could drop 10 feet through the rushing rapids beneath. So you pay the shootsman who was contracted by the bridge to guide you through the piers. And it was really quite dangerous.
So, the bridge has a number of maintenance problems: the first is the starlings the mason repairs. The second is until the mid 18th century the bridge was covered in housing just like Ponte Vecchio in Florence as a proper living bridge the housing was also in a state of disrepair and some of it owned by the bridge and some of it owned privately. So the bridge tried to take over the property that isn't theirs and then get rid of the housing that isn't working, it's falling into disrepair over this period.
And there's a guy called Mark Leighton who's written a brilliant thesis at the University of Leicester all about how the bridge masters and the City of London get rid of the housing in the mid 18th century.
But essentially the bridge is the only crossing from side to side, from north to south or vice versa until 1750. There isn't another way to cross the Thames. There was a little wooden bridge up in Putney in 1729. London Bridge it's got all of the infrastructure of London basically. And so it's hugely congested and falling apart. So, the maintenance bills are are huge.
Oh yeah as well. So as well as the starlings you then have water wheels which are basically bringing the water from the New River Company and the Thames to give water to the city. So those are also in operation, these whole teams of little engineers looking after the water wheels. So it's a really busy bridge it's got people scrambling over it all the time looking after it, not before the shootsman or anybody else doing any work on it and those people were paid not very much.
The master craftsmen were paid for their contract and got a really good rate for looking after the contract, and then they hired others piecemeal so they'd hire well-known carpenters or masons. But they'd never have regular days or regular work and then the labourers were paid by the tide.
So at high tide you could work on the bridge or you could work on the upper bits of the bridge if you were in a boat; at low tide you could access all those damaged starlings and piers. So at low tide they worked in boats and that meant that in the winter you might only get four tides in the week depending on when the tide and the light coincided, in the summer you could maybe get 11 and then when they didn't need any work done you wouldn't get any tides at all.
So, there were quite a number of people. It varied from teams of 12 to teams of 80 or so who were employed in this fashion in a piecemeal just waiting for a little sort of bit of peace work on London Bridge. So, it's an interesting bit of contact with the sort of materiality of the world as well, everything was literally ruled by when the water came in.
Petersen: Right. And since it's such a long period of time, I suppose you can get a decent time series of that change in the wages over that period.
Stephenson: Yes, from a labour economist point of view, one of the fascinating things about the 18th century is this persistence of rates, particularly for labourers, it's a very monopsonistic market it's a classic monopsonistic market. It's a wage posting. One where employers basically will see who will come at this set wage and what happens is they don't change the wage.
The fluctuation happens around the number of days worked. So people don't turn up, or don't get work when there is less to do. The number of days fall away and when there is high demand, an upward-sloping curve, the number of days go up for everybody.
But a transaction cost analysis would suggest that the 18th century employer understood the costs of such information very well indeed because they weren't going to have any asymmetry of information. They were going to post ‘this is what you get,’ particularly the unskilled hand and the time or the amount of work that you got was how the fluctuations and the dispersion occurred.
So there's a lot more work to be done on that because nobody's really ever looked at this kind of market in those modern terms, understanding it as monopsonistic or having search or information costs.
And it's only with these levels of micro data that we can begin to understand that it might have worked like the labour market we know. Until about 20 years ago people thought---until much more recently actually, the last paper I can see about this is in 2007 by Leonard Schwartz---that essentially before 1840 it's a market dominated by custom not by market forces. But on a micro analysis it looks very much like there are just the kind of market forces at play that we understand today. So, wage posting at the lower level, a little bit of wage bargaining at the skills level, and supply and demand do actually equilibrate but not through the rate, through the number of days worked, which of course brings about the income.
Petersen: So, the third construction project you discuss is the Westminster Bridge, which I suppose is that that second bridge you mentioned earlier.
Stephenson: Yes, the second bridge, the cross rail of the 18th century.
Petersen: Is that interesting from an economic history point of view, we have a lot of data from that?
Stephenson: You get less data because I don't have anybody's nice little book saying who came in and on which day, so I don't have the number of days' work for Westminster Bridge.
The interesting thing about Westminster Bridge is the different kinds of contract. Everybody, they were making contracts for hundreds of thousands of pounds with the masons and engineers and they also had a contract with a guy who had a horse and three piles for 27 pounds for the year. So, you've got this variation in value or risk from a financial point of view which is quite dramatic.
But the key thing is that at Westminster Bridge you find the tide and the day model as well. So a much smaller number of days than you would expect that are actually billed to the institution, but this means of paying by the tide, which protects productivity from an employer's point of view. So that also occurs at Westminster Bridge.
And what you find is that people are doing quite advanced and quite dangerous work, but without the danger money. They were given gin instead. So they sank caissons, this is one of the earliest uses of caissons designed to create the piers. So these things are experimental to say the least, and they put people in diving gear into the caissons and it must have been terrifying, you know, what if the stuff gave way and they went under the Thames. In February, because that's the time you want to be in the Thames! You know, in 18th century diving gear. And got them to work on the masonry or on the carpentry on the bed of the river for the same rate as you could be having quite a nice comfy time carving out something simple, or doing some basic maintenance work on a couple of windows on some bridge houses. So, yes very dangerous work. There seemed to be a lot of skill available, ready to do that work at those kinds of rates.
Petersen: So, where do you see this research program going in the future?
Stephenson: There's obviously an issue about the rate of welfare, the real wage and welfare in the 18th century and to be honest if we're going to make a serious contribution to that, we need to start looking at people who aren't builders.
I've started a project with the Cambridge Group for the History of Population and Social Structure, where I spent a year before I went to Oxford, on London occupations. Because that Cambridge group, they are the masters of working on occupational structure in the long run in England and we are sampling institutions that bought goods and services widely. And the kind of bills and the kind of businesses that they deal with to understand what sort of people were employed where. So, to try and get some welfare and some wage data beyond builders that we can normalize and use properly.
I think the second direction for this research is to understand how labour markets worked. Was there such a thing as custom? Because one of the old things we believe about the Industrial Revolution, and this idea doesn't really stand up anymore, but it's something that's still emotionally alluring for a lot of people, we see the Industrial Revolution as that sort of capitalism thing and our version of capitalism got going.
But if people already understood transaction cost economics, and Christopher Wren writes like Oliver Williamson sometimes, then maybe the market didn't start then, maybe they already had a view of the market. And there are some organizational things that we need to be looking at from that point of view.
Essentially the 18th century will always be interesting because it is a free market. It is unregulated, there's no corporation tax and the finance is not state controlled at all. This is before the gold standard, this is before states get interested in managing money in a big way. There is monetary policy but it's not in the same way we conceive it now. And so labour and capital have a relationship that is unencumbered by the state, by government, by regulation.
So what is the outcome of that? Was it a race to the bottom, was there any equilibrium, what happened? So, there's a contribution to be made to studying that as a sort of a history of ideas thing as well. It's hugely rich but those are broadly the three things that are on my agenda right now.
Petersen: My guest today has been Judy Stephenson. Judy thanks for being a part of Economics Detective Radio.
Stephenson: Thank you very much. I very much enjoyed talking to you.
Fri, 16 December 2016
What follows is an edited transcript of my conversation with Samuel Gross.
Petersen: You're listening to Economics Detective Radio. My guest today is Samuel Gross of the University of Michigan Law School. Sam, welcome to Economics Detective Radio.
Gross: Great to be here.
Petersen: So our topic for today is criminal justice, in particular, we're going to be looking at the issue of wrongful conviction. Dr. Gross was part of the establishment of the National Registry of Exonerations which has provided valuable data in this area. So let's start by talking about the registry. What is it? How was it developed? And what was your part in it?
Gross: I'm the founder of the registry. It was created because after doing work on false convictions and exonerations for half a dozen years it became clear that the only way to get any sort of systematic information on exonerations that have occurred in the United States would be to put together the wherewithal to collect that information directly because nobody else was doing it. There's no official system for gathering information on exonerations or for that matter a single legal definition of what is an exoneration. And from there this project just took off on its own and became what's now a lasting institution that's in the process of handing over to other people to run.
Petersen: Okay, so let's talk about the how an exoneration is defined in the registry. It's a case where someone has been convicted and then later that conviction has been overturned and presumably not just on a technicality but because the person was actually innocent---somehow managed to prove their innocence. Is that correct?
Gross: Something like that. The thing that we're interested in studying is false convictions. Convictions of people who are actually innocent. The problem is there's no way to know when that's happened directly because in cases where by assumption people have made errors, the juries and prosecutors and judges who considered the cases have made errors in deciding who is guilty and who is innocent. It would be foolhardy and presumptuous and not particularly accurate for us to believe that we're going to be better at deciding after the facts which cases involve those sorts of errors and which don't.
So, what we've done instead---instead of trying to make any judgment of our own about guilt or innocence based on whatever information we can collect from second or third or fourth hand sources---what we do instead is to consider cases, to classify cases of exonerations if they meet the following criteria: First, the defendant has to have been convicted of a crime, not just charged, the conviction has to be completely removed. That is, at the end of the process the defendant has to have no legal consequences from the original conviction which means the conviction was entirely overturned by dismissal, or by a pardon, or in a small number of cases---or by the acquittal over retrial---or in a small number of cases by one of the few procedures that are available for people who are exonerated posthumously, after they're dead. And then in addition, the process that led to that result has to include substantial evidence of innocence that was not available at the time of the original conviction.
If the case meets those criteria we include, if it doesn't we don't. Our belief is---and it's a hypothesis---is that this produces a conservative classification for actual innocence, for errors in determining guilt at trial. We know of quite a few cases of people who are very likely to be innocent who are not included by these criteria in particular. Quite a few cases of people who reach that result having the conviction entirely removed from the record without the production of new evidence of innocence. For example, cases in which the conviction was reversed and then they are dismissed want to appeal because there was insufficient evidence to convict at trial unless some other evidence comes up before the conviction is reversed don't generally count.
And our hypothesis is that we include a very small number of cases of guilty people who did meet these criteria. Although there are no doubt some, it's possible, but the process that it takes to get convictions reversed and dismissed in this manner in the United States is very difficult and as a result as far as we can tell there are not very many misclassifications of people who did prison based on the underlying crimes but meet the criteria for exoneration. That is the best we can do.
Petersen: One thing I learned from your research is that the average time it takes a wrongfully convicted person to be exonerated is 10.1 years. So it seems like the system is stacked against it.
Gross: It changes depending on the mix of exonerations we have. In the moment it's probably gone down a little bit, but that's about right. Ten years is close to the average
Petersen: Looking through the registry I noticed some interesting patterns that I wonder if you'd like to comment on. So for instance, I was surprised by how few of these exonerations were about DNA evidence. It seems like it's actually more common for a witness to recant their testimony or for other non-DNA things, but I guess looking at media or just my own intuition I would have thought that DNA would be the big factor in a lot of these.
Gross: And that's a very common misconception. I think most Americans think that exoneration is the second word of a two-word phrase that begins with the letters DNA. And that's of course, as you know, not true. DNA exonerations have always been a minority and in the past 10-15 years, they are an increasingly small minority.
The number of DNA exonerations has been relatively steady over the past 10 years, about 20 a year, and the number of not DNA exonerations has been going up rapidly. The basic reason behind this is the DNA, which can be very telling and provide extraordinary strong evidence of guilt or of innocence, is only valuable in a small range of cases. It depends on having biological trace evidence that identifies a particular person as the person who committed a crime.
That's relatively straightforward in the case of sexual assaults---rapes in particular---where the trace evidence that's left is very hard to explain except as a consequence of the crime. So if you recover semen from the body of a rape victim and it's identified by DNA as coming from a particular person and that person is not a consensual sexual partner of the victim, then you have the rapist. Assuming that rape really did occur. But if it's a question of identity, which is the case in most of the rape exonerations we know about, that tells you both who the person is if you can identify the profile---and in the case of exonerations who it is not---because DNA comes asymptotically close to being a unique identifier in that type of DNA evidence.
But that's rape cases. Some other violent crimes produce DNA evidence that is as valuable or nearly as valuable. Usually murder cases in which there's blood evidence sometimes, other types of biological trace evidence, perspiration, epithelial cells from the skin and other bodily fluids and that can identify people under circumstances where you can determine from other evidence that the biological sample that was retrieved could only have come from the person who committed the crime.
But most crimes don't have that. One of the things that's clear in the registry, for example, is that there are almost certainly many many cases of false convictions in robberies that are not exonerated because they don't have the type of evidence that's available in rape cases. And in many rape cases there aren't either, but both robberies and rape cases that resolved in exonerations overwhelmingly are cases in which the suspect was misidentified by the victim or by sometimes more than one victim. There are three to five times as many robbery cases of this sort as rape cases---maybe more than that---and they are typically cases in which errors are more likely because the victims may only get a sidelong glance at the robber whereas rape cases almost by definition require much more close contact with the perpetrator.
Nonetheless, despite the fact of the numbers suggest we have many more exonerations in robbery cases. Rape exonerations outnumber robbery exonerations by about 3-5:1. And the reason of that is because the rape cases can be exonerated by DNA but the robbery cases can't because you don't have DNA for the type of conduct that's involved in robberies. You don't have DNA that you can retrieve from threats that are made or guns that are waved or even guns that are used. And as result, those cases are not exonerated at all. Now given the huge disproportion between the cases in which DNA is valuable, and cases in which DNA is never going to play any role, it's not surprising that the great majority of exonerations don't involve DNA. But the availability of DNA in the small number of cases---comparatively small number of cases---in which it is valuable does make those cases ones in which the possibility of correcting a terrible mistake is considerably harder.
Petersen: So, you mentioned the increase in recent years in non-DNA exonerations. I was looking at the data and in 2011 there were only 73 exonerations in the Registry, but in 2015 it had more than doubled to 157. So, I wonder what's changed in recent years to make that number increase so much?
Gross: It is a general change and then there are particular---couple particular strands that stand out. The general change is that the resources that are devoted to reinvestigating cases where defendants were convicted and there are now doubts about the accuracy of the convictions are increasing year by year. And the willingness of everybody involved, the criminal defense attorneys but also more importantly prosecutors, police officers, and judges to consider the possibility that someone who was convicted is innocent has grown greatly.
They have come to recognize that mistakes happen and the more exonerations occur the more people realize that this isn't just a once-in-a-lifetime event. And that's I think the force that is behind all of the specific changes that occur.
The two strands that have made the most difference in these numbers are---in the last 10 years, I think---are, in the past four or five a proliferation of Conviction Integrity Units across the country. These are specialized units within prosecutors' offices that focus on issues having to do with erroneous convictions. And for the most part, the ones that are most effective, look at cases within the jurisdiction in which there is a possibility that the wrong person is convicted and they work to re-investigate and sometimes exonerate the people involved. They have contributed an increasing proportion of the cases that we see. And to the extent of this becomes more widespread they might someday be a majority of all exonerations.
There are something like 25,000 local prosecutorial offices in the United States. I don't have the exact count now, I will in several weeks but my guess is that by now we have something like 30 Conviction Integrity Units around the country which represents a larger proportion of the population than that would suggest because those are some of the most populous counties, but it's still a minority of all cases where the local prosecutor has any organized interest in the issue. And then there is a particular pattern that came up in one county---Harris County, Texas, where Houston is located---that contributed, I think it was 40 some exonerations last year and 40 some so far this year and 30 several years before that. And that's a back-log of cases of defendants in Harris County who pled guilty to possession of drugs and then after they pled guilty the Houston police criminal lab tested the substances they received from them to determine whether they, in fact, contained drugs and found that the material that was the basis for the conviction included no controlled substances whatever.
We ran into a number of cases as that over the years but starting in 2014 the attorney who runs the Conviction Integrity Unit in the Harris County district attorney's office noticed there was a whole bunch of these cases and they were being handled very haphazardly and put together a program to identify them all, to clear the backlog and to set up procedures for getting the defendants exonerated and they're still working through that. The thing that's interesting about those cases is that, that procedure testing drugs that were seized from people, the supposed drugs that were seized from suspects after the defendants have already pled guilty to something that as far as we know doesn't happen any place else in the country or didn't---now I think a couple of other jurisdictions have begun to do it at least occasionally---but it's just that there may out there be thousands of cases a year of defendants who pled guilty to possession of drugs when in fact they were not in possession of any illegal drugs. Although they could be exonerated by quite a simple process, it doesn't take an elaborate investigation just running the drugs through the police lab, which could be done. Except in Harris County, until the last year or two nobody's ever done that.
Petersen: That's so strange. Were they caught with a little baggie of oregano or powdered sugar and just everyone assumed it was drugs?
Gross: It's a whole lot of different things but it's a good question. In some cases, they were arrested for possession of pills that were identified by the police officer as likely or actually being a controlled substance---often Xanax---and then tested by the lab and found not to be Xanax. Sometimes ibuprofen, sometimes some other over-the-counter medications.
In some cases, they had the smallest amount of white powder. One woman who ended up in the paper had white powder on her face because she had the habit of eating flour, which some people do, eating flour mixed with water and was left with some white powder around her mouth. In cases like that, where small amounts of white powder or something else were tested, they were subjected to field tests for drugs which have become notorious in the past year or two because they are so unreliable.
Field tests for drugs are not admissible evidence in court to show that the substance involved was a controlled substance but they're considered good enough to perform the arrest. Then what happens is the defendants who were arrested on these charges show up in court three days later and if they can't make bail, which seems to be true in basically all the cases that we know about---perhaps with a few exceptions---then they're given a choice that amounts to this: Plead guilty today and you'll get some perhaps suspended sentence and go home immediately, or another week and you'll be home in three days, or something like that, or a short-end by Harris County standards where drugs sentences are three weeks or two months or something like that. Or plead not guilty and then you'll be held in jail because you can't make bail for months, after which you'll go to trial and if you are convicted, and obviously you might be because some cop and some test said you were in possession of drugs, then you'll get perhaps years in state prison.
No doubt many people refuse to plead guilty in that situation because they're innocent. But some who are innocent do plead guilty and those are the ones we find out about.
Petersen: Right. So that process of being prosecuted, even when you're innocent, can sometimes be more costly than simply pleading guilty. Especially in the small cases.
Gross: If it's a low-level charge and you are held in pretrial detention, the process can be much more costly than a conviction after trial. In Harris County you could spend six months in jail or longer. So, I've heard stories of people who spend a year or more in jail waiting for trial. If you can make bail, then that's still a heavy cost. Having a trial hanging over your head and having to come back to court repeatedly is not a walk in the park. But if you're held in jail for that period it's an extraordinary cost. It obviously disrupts your life and it may tear your family apart, you would lose your job, you may not be able to get other employment, and so on.
Petersen: So one of the things we've learned in recent years is just how easy it is for police to get people to falsely confess, or to falsely accuse someone else of a crime. In fact, I think the number was 12% of the cases in the registry were false confessions and I believe it was a majority involved witness making false accusation or committing some kind of perjury. Why do people confess to things they didn't do, first?
Gross: Well, those are two different issues false confessions and perjury or other false accusations. Although there is sometimes an overlap. I don't think I agree with the statement that it's easy to get false confessions. In some cases, no doubt it is, but the false confessions that we know about are overwhelmingly in murder cases. And as far as we can know---but the information we have is not that good---they appear to be the result of long interrogations. Long would be the ones that really stand out, sometimes they take place over days, two, three days or longer with the defendant sometimes being questioned in relays by different officers.
But even interrogations of three, four hours or five hours---which is eternity if you're being questioned unrelentingly by police officers---are fairly uncommon. This costs the police quite a bit. It involves usually more than one officer over that period of time. Certainly one officer taking off that whole period of time and a fair amount of preparation. So it's not done very much, except in homicides, and it does happen in other cases but I think two-thirds or so of the false confessions that we know about are homicides cases, maybe more than that.
What's surprising to many people---and those working in the area have gotten used to it, but still somehow surprising---is not that it's easy to get people to falsely confess, but that people will falsely confess to murders that they didn't commit, that they had nothing to do with. It seems like such an unbelievably strange and self-destructive thing to do but it happens and it happens time and time again.
And what we see in these cases and what other researchers have shown, is that it's much more likely if the suspect is in one of two general very vulnerable categories: if the suspect is a teenager, or the suspect has some type of mental disability, is intellectually disabled and mentally ill. Those two groups are, as far as we can tell, much more likely to falsely confess than other people who are subject to these interrogations. They are very far less likely to be able to resist the pressures, they're more likely to respond to authorities or more likely to believe the promises and threats that are implied if not directly made in the process of interrogation. They're more likely to become hopeless, and they're more likely to---much more likely---to not grasp the seriousness of what's going on.
One of the most common things that people hear when they talk to a suspect who falsely confessed later on and ask why did you do it is "I confessed because I wanted to go home. I just told them what they wanted to hear so they'd let me go." Which of course does not happen or "I confessed because I just couldn't stand it anymore so I told them what they wanted to hear but I didn't think anybody would believe it. How can anybody take this seriously?" And they think that because they're there and they've been experiencing this onslaught going on for a while and they imagine that anybody knows what happened would say "why would anybody take seriously what somebody says after hours of being badgered and humiliated and lied to?" But of course, the jurors and the judges hear the confession at the end typically don't know what happened but they have a confession that has the defendant's name at the bottom.
Petersen: Right. So the other thing I was asking about is the false accusation. A lot of this comes in child sex abuse cases. There's a pattern in the data.
Gross: Yes, that's correct. Child sex abuse cases are overwhelmingly the cases in which there is no DNA or other physical evidence of the sex abuse. They are typically made on the accusations. The crimes are typically made anywhere from weeks to years after they occurred, way too late to have any kind of physical evidence.
And that as far as we can tell is accurate as well as inaccurate child sex abuse accusations. The children who were victimized by adults many, perhaps the majority, probably never report the abuse at all. And those who do, everybody understands, will not necessarily do it right away. Sometimes they have to be encouraged, but that means that false accusations are very hard to detect because what do you have? You have the nine-year-old girl who says that her stepfather molested her in some way, repeatedly over the period of a year. So ending a year or more before she ever tells about this. Nobody can figure out exactly what happened or what date this occurred, where people were, whether anybody else could have observed it, etc. The details of a crime that occurred in that context are essentially impossible to recover. To defend against it using evidence of timing, or the presence of other people, is generally impossible.
It becomes a contest of credibility and that means that if somebody is motivated to lie in the situation---and we see that in the cases that result in exoneration---the defendant may well be convicted and that will be the only evidence. And there are many of these cases in the registry. And they seem to often include cases which we don't find out about until years after the exoneration took place.
Often cases got little attention in the time for exoneration. I don't think we can come up with anything like an estimate of how come, but we do know that they're not rare.
Petersen: Moving to a different topic, you co-authored a paper published in 2014 on the rate of false convictions among death row inmates. I should say, I first heard of this study when someone quoted a statistic from it, the statistic that at least 4.1% of death row inmates are innocent, and my first thought when I heard that was that it couldn't possibly be right because nobody should have the tools to measure such a thing. So I read the paper and I was impressed with the method. So could you talk a little bit about the methods you used to get to that statistic?
Gross: Well, I have to agree with your initial intuition because having worked in this area for decades, 20 years ago I would have said this is not something that can be estimated. The proportion of false convictions can't be estimated because false convictions, are by their very nature, unobserved and for the most part unobservable.
You can't just have to find whether somebody is innocent or we wouldn't have false convictions. A lot of things are unobserved. We don't know the rate, the proportion of prisoners in the prison who have been exposed to tuberculosis, but we can test that if we want to. We can test a sample. But this isn't anything you can test. So the question is "Is there any way to come up with an estimate?" And in general, it's impossible because you don't have anything like the post-conviction information that will be necessary to come up with a clue as to how frequently innocent people are convicted and for what it's worth in the United States our background criminal justice statistics on convictions in general, are so bad that a comparison of people who are convicted is nearly impossible to define.
Death sentences in the United States are just different. They're different in two ways: First, we have through the Bureau of Justice Statistics a database that tracks everybody sentenced to death in the United States with reasonable precision from the time they're sentenced until the time they're removed from death-row by execution, or by death from natural causes, or by exoneration or by the most common means which is being re-sentenced to life in prison. And second we have an extraordinarily high rate of exonerations in death-sentence cases. Hundreds of times higher than for other crimes. And the reason for that is---depending on what you're comparing it to---perhaps only 10 or 20 times higher than other murders but much higher than other felonies in general and vastly higher than misdemeanors.
And the reason for that is that someone who is under sentence of death almost always will have access to resources to reconsider the possibility that he or she is innocent that are simply unavailable to almost everybody else who has been convicted of a serious crime in the United States. They're going to be represented by attorneys with few exceptions from the time they're convicted until the time they leave death-row, their cases are all subject to review in private courts which is not generally true for other people sentenced in the United States and in almost all cases there are multiple levels of review.
The resources that are available for attorneys are much greater and probably as important as any of that, the legal system itself---and judges in particular are much more interested in considering the possibility that people who might be executed could be innocent than they are in considering the innocence of anybody else including defendants who are sent to prison without the possibility of parole---and are much more open to reconsidering cases where new evidence of innocence comes up.
So the net result is a rate of exoneration that is vastly higher than any other type of case. Which means that if you just look at the rate of exoneration and compare it to the overall number of death sentences that occurred in the time period where the convictions which produces exoneration occurred you already get quite a high number, about 2.3% if you limit yourself to cases that are old enough so that by then anybody who will be exonerated probably was exonerated. That's a paper that one of my co-authors---Barbara O'Brien---and I published several years before the one you are referring to.
So that already gets you to a number that is surprisingly high but it's still going to be an underestimate. An underestimate because that also takes into account the many cases of defendants who are sentenced to death who are innocent, who have not been exonerated. The paper you mentioned tends to deal with that by focusing on one of the features of the process of exoneration and review of the cases of defendants who were sentenced to death and that is something I already mentioned.
If you're sentenced to death in the United States chances are you will never be executed. What will probably happen is that eventually, by one means or another, your sentence will be reduced from death to life in prison and you will be taken off death row and reassign to the general prison population and then you will die in prison as too many defendants who were convicted of murder and sentenced to life in prison and not sentenced to death. When that happens, the pressure to make sure that no innocent person is executed, which is the backbone motivation for the extraordinary high level of exonerations in death cases, is removed. The defendant is no longer under threat of death and the extraordinary resources and attention that that defendant gets go away. And as a result, what you see is the death penalty exonerations that occurred are overwhelmingly in cases of people who still remain under sentence of death and are on death row. And then once the threat of execution is removed, the rate of death sentencing drops back, as far as we can tell, basically to the same rate as other murder cases.
So the question for us was is there a way to estimate from the pattern of the cases that we know about---in particular, their histories as they wend their way through this process---what the rate of exoneration would be if defendants who are sentenced to death remain under threat of death indefinitely? That is, subject to the type of searching investigation and reconsideration that's available to defendants who might still be executed.
And there is a technique for doing that, which you're very likely familiar with, and that's survival analysis. Figuring out how to do it is a somewhat complicated process but we went eventually through and that produced the estimate that you see, which is 4.1%. Obviously there's a great level of uncertainty attached to that. But it is a legitimate estimate of the rate of false convictions in that context, given the assumption the underlying rate of exonerations is a decent measure of innocence. And that requires a sensitivity analysis, which we also go through in the paper.
And that leads to, I think, a solid conclusion that this is a conservative estimate of the rate of innocence among defendants who were sentenced to death. It's the lower bound of the point estimate, but it really means something between 3.2 to 5 point something percent. It's hard to know what it means, but it's somewhere in that range of 1 in 20 to 1 in 30. Which, I have to say, I was surprised how high it was.
Petersen: So where do you see this research program going in the future?
Gross: Well, unfortunately I cannot see duplicating that type of estimate for any other categories cases because we do not have that information. We don't have the background information on the cases themselves, which is necessary to do the survival analysis. You actually have to know how many cases survived, what the trajectory of each case was through the legal system and you don't have a glimmer of that except for capital cases. And second, the exoneration rate is much lower. So you can't make the leap from the exoneration rate to the false conviction rate. It would be much more tenuous.
Our estimate at the end amounts to saying that we detected something like 40% of the innocent people who were sentenced to death in this period in the United States. The rest, the great majority of them ended up in prison---under sentence of life in prison---without the possibility of parole and will probably remain in prison until they die.
Some number of them were no doubt executed. Although it follows from the same logic that leads to this estimate, that that will not be anything like 4%. That of the 12,000 of people who've been executed, if they were being executed in proportion to the number of people who were convicted who were innocent, it would be like 50 or something like that. But since so much of the process is geared to avoiding executing innocent people and that produces such a high exoneration rate, my guess would be much lower than that. It would probably translated into execution of probably 10-20-25 of innocent people over the past 30 some years. But that's it.
Again we don't know, we can't say which cases they are. But can we do the same thing for robberies, or kidnappings, or for that matters non-capital murders? I can't imagine that.
Petersen: My guest today has been Sam Gross. Sam thanks for being part of Economics Detective Radio.
Gross: My pleasure. Take care.
Fri, 9 December 2016
What follows is an edited transcript of my conversation with Otto Lehto.
Petersen: You're listening to Economics Detective Radio. My guest today is Otto Lehto of King's College London. He is formerly the chair of Finland's Basic Income Network. Otto, welcome to Economics Detective Radio.
Lehto: Oh it's my pleasure to be here.
Petersen: So our topic for today is the basic income guarantee. Otto, you approach this idea from the perspective of political philosophy, so let's start by discussing that. How about we start by talking about two of the major figures in political philosophy: John Rawls and Robert Nozick. What do each of them have to say about the welfare state and where do your views diverge from theirs?
Lehto: That is a good point to start indeed, although it is I think a bit lamentable that we have to start from those two figures because they have dominated the discussion so much during the last 50 years. In fact, it's very hard to have a conversation outside the boundaries set by those two figures, but they're both geniuses. They set the stage for the discussion, certainly in philosophy but also in public policy in many respects.
So, let's start with John Rawls. John Rawls really was a towering figure in Harvard, really starting from the 60's and throughout the 70's. He wrote this book, A Theory of Justice, which is considered one of the really truly great books in political philosophy that revolutionized the way we think about these subjects. But the short version of his theory, which is very influential even up to this day, is that people in societies should look at the framework of living with each other as a cooperative game where we all try to sort of not only maximize our own position but also to make the whole game fair for everybody. And so he called his theory Justice as Fairness, where people are entitled to a certain respect and autonomy, certain liberties as members of the democratic community where they can pursue their own ends. But they're also entitled to a redistributive scheme if they happen to be among the worst-off people in the society. They are entitled to redistributive transfers.
This framework sounds very familiar and indeed it should because it reflects the social democratic reality in which most Western societies operate. And even in later years he said that actually his philosophy, even though it starts from first principles and proceeds from there, is actually meant to be a philosophical justification of the intuitions that people in Western democracies---liberal democracies---have.
So, you combine liberal ideas of individual freedom with these notions of the welfare state and so on. So that was the foundation of Rawls' system.
So that's Rawls' system but Nozick came along and he found a place for himself in the same institution, that is Harvard, and he wrote a critique---a respectful critique---but a very thorough and deep critique of Rawls' theory. And he ended up justifying a minimal state that libertarians are very fond of. And he effectively said that no, people should just be seen as individuals who have some fundamental rights---he calls them side constrains---that people have a certain respect that they are owed by other people and it is very wrong for people to violate their personal boundaries and this includes the State.
The state has actually no right to violate the sort of inviolable right to property rights that individuals have. So every form of taxation, that features very prominently even in Rawls' system, is theft. So, that is of course a very prominent theme in libertarianism. So his book---which by the way is really brilliant philosophically, it's not only just a standard justification of libertarianism but it's actually one of the great books in philosophy because it's so rich and powerful and full of interesting ideas and strange examples and brilliant footnotes and all that---but that lay down the other side. And so the debate in intellectual philosophy and history in the last 50 years or so has been largely dominated by these two figures: Rawls' Theory of Justice on the one hand, a justification of social democracy with a liberal bent, and then on the other hand Nozick's Anarchy, State and Utopia, which is a justification of libertarian taxation-is-theft ideology. So that is the framework in which we find ourselves.
Petersen: So there are these two competing extremes. You quote John Tomasi's critique of both of them. Would you like to summarize that for me?
Lehto: Yes. John Tomasi wrote a wonderful book in 2012 called Free Market Fairness where he actually tries to combine these two perspectives. And he says that actually there's a whole tradition that we're forgetting here when we focus only on these two---as you put it, they are both at extremes---although at least for Rawls himself, he's often considered a centrist. But in many ways, he represents this kind of---from a perspective that Tomasi points out---the perspective of classical liberalism even though the Rawlsian center-left position, he's actually seen going fundamentally wrong in many ways, even though that is the unquestionable framework in which people today operate.
And I should say, when I say that Rawls and Nozick laid a framework, it's not as if there is 50% on one side and 50% on the other side. Perhaps in politics, like the left-wing and right-wing ideologies have maybe about 50% on each side depending on the circumstances. But in philosophy certainly, Rawls has been the one that dominated the discussion and there are actually very few Nozickians around.
But Tomasi points out that even with this seemingly very credible and too wonderful system that Rawls lays out, there is very little attention paid to issues like individual freedom especially in the domain of economy. And the lack of respect for people's freedom of choices in economic matters is actually a major shortcoming in Rawls' system. And this is exactly what Tomasi points out and from the perspective of classical liberalism which he raises to the standard of something that we should actually take more seriously than we have today. He points out that actually economic liberty is something we should insert back into the conversation in a serious way without however on the other side falling down the assumption that Nozick makes---and a lot of libertarians make---that the only justification for all economic liberty necessarily leads to a justification for the night watchman state or the minimal state of libertarianism where there is no role for government to provide public services and all that.
And so this false dichotomy that Rawls and Nozick have put out has sort of made it difficult for people like myself and Tomasi and Matt Zwolinsky and people who consider themselves followers of the legacy of classical liberalism to lay out the more complicated, but I think more interesting, case for a system where robust economic liberties are combined with certain welfare state elements. Certain elements of taking seriously the power of the state to actually increase the real opportunities of people rather than just being a system of theft as Nozick calls it.
Petersen: So that's where something like the basic income guarantee comes in. Can you summarize what that is and how is that different from the welfare states most countries currently have?
Lehto: Right. Basic income guarantee, first of all, is defined as a regular payment to all citizens or residents of a political community that is given uniformly to all citizens. All people get the same amount and people get it without bureaucratic discretion. So it is given automatically or almost automatically to all people either in the form of a direct cash transfer to their bank account or in the form of a tax break system as in the form of negative income tax which is actually a form of basic income.
So this system is supposed to, and it is a way, to replace the bureaucratic complexity and the nightmarish disrespect for human autonomy and human freedom that lies in the center of the current welfare state system in my opinion and certainly in the opinion of Tomasi and other people who I'm referring to. So the basic income guarantee is superior to the current system and it differs from the current system in the sense that it actually operates under the principle that we shouldn't use the state to guarantee specific favors to specific people, we shouldn't use the state as a one-upmanship mechanism whereby one group of recipients carries for the favor of bureaucracies, tries to---and in a way infiltrate---the mechanisms of the state to redistribute money and resources to themselves or to groups that they favor against the interests and desires of other groups because this leads to a spiral of negative-sum game in the political economy. And I think welfare states today in this sense have become victim to this overzealous one-upmanship of special interest group politics and basic income is a way to overcome this problem.
Petersen: So the basic income guarantee, is it really a break from business as usual? It seems like it's a marginal improvement on the system we have now, but I guess you're suggesting that the system we have now encourages a lot of rent seeking, it has a lot of payments to different groups, it's needlessly complex. I could list some other problems with it. There are the so-called welfare cliffs where poor people face implicit marginal tax rates sometimes of a thousand percent, or some absurdly high amount because their benefits are clawed back when they earn a little more income. So there seems like there's a good economic justification for basic income. Is your work focused on the classical liberal philosophical justification for having a hands-off welfare state?
Lehto: Yes, in a way. The fundamental debate is truly between these two perspectives of whether it's a pragmatic justification for reform towards a slightly saner and slightly more useful and purposeful and beneficial system, or on the other hand, is it a requirement of justice that we have something like a basic income guarantee. And I think that really the truth is somewhere in between.
First of all, I think it certainly is a pragmatic improvement over the current system but I should point out already at this point that when I'm advocating for basic income I'm not advocating for basic income without demanding widespread reforms in other areas of life in the welfare state. I am indeed calling for massive restructuring of many of the mechanisms of the welfare state partially just to accommodate for the fact that we are taking basic income as the policy paradigm that we're trying to implement. Because if we take that as the policy paradigm, then we necessarily must reform the existing bureaucracies, tax system just to accommodate for the fact that we are taking this new system into effect.
In addition to this, I think that the whole framework of regulations, the whole framework of massive interventions into the economy, into the private life of citizens have to be addressed as serious violations of the capacity of the welfare state to truly increase the welfare of its people. Because my opinion is that the welfare state has failed because it has failed to address the proper means to achieve its own ends that it claims to have. Use of improper means to achieve its ends is the reason why the welfare state is failing so miserably everywhere in the world today. That it's claiming to be for the welfare of its citizens, but if you look at it in terms of its overall effect in many ways it fails.
Petersen: So, when I think of the policies that I'd like to see replaced by a basic income guarantee they're not just strictly welfare transfers. There's a theorem in economics called the Atkinson-Stiglitz theorem. It says that when you have an optimally designed progressive income tax scheme, basic income with a progressive income tax would be something like that, then it doesn't make sense to have additional programs designed to redistribute. And some of the programs that I think are basically focused on redistribution are things like protecting taxi drivers from competition from companies like Uber and Lyft, or a lot of the interventions into medicine are designed to make sure that people who get sick don't also become poor. And of course, if you had something like a basic income, every taxi driver could lose his job, he wouldn't fall below that minimum level. And so could at least in principle---if we were going to make sort of an ideal political bargain---a basic income guarantee would come with a lot of free market reforms ideally. Is that basically a big part of the reason why so many libertarians---such as Milton Friedman and Friedrich Hayek---have supported versions of a basic income?
Lehto: Well yes indeed, it has the feature of being compatible with a total abolition of the rest of the welfare state, or major portions of the welfare state. And in fact people like Charles Murray have recently proposed exactly that, a replacement of the welfare state by the means of a basic income given to all citizens as the second-best option to a complete free-market society. And people like Hayek and Friedman were also of the opinion that the majority of those transfers could be replaced.
So the thing with money is that money is a universal means of exchange and the uses of money and the need for money are as varied as people and situations. And when we think of basic income we don't think of it in terms of being for a particular purpose or for particular people or for particular circumstances unlike the current measures. And so it has the virtue---and perhaps the vice depending on your point of view---of being this universal situation, a neutral ground. And so indeed we can come up with hundreds of scenarios where a basic income could be useful for people. Obviously, some of those are covered by the current redistributive schemes within which by the way I would include things like farm subsidies, many forms of corporate welfare and so on.
So basic income has the virtue and vice of being neutral as regards purposes and situations. The only thing really is that if you don't have any other sources of income then you will get a basic income without having to beg for it from anybody either in the government or in the world of charity for example. So, yes indeed, people who are forced out of work to circumstances---whatever those circumstances happen to be---are able to survive, the people who are forced out of the labor market entirely for a reason---one reason or another---people who have temporary or permanent conditions that affect their capacity to find work will be covered up to this level, and people who perhaps want to take some time off to take care of their family, people who want to take some time off to study, to plan ahead, to perhaps think about starting a new company, they have some ideas but they don't have the means of funding yet, that allows people to focus on doing what they think is best for them at the moment. So it has almost an infinity of purposes precisely because there is an infinity of human beings and human desires that in a pure realistic society will have to be taken into account. And a welfare state that tries to measure what people truly need, or what circumstances need to be taken care of, fails precisely because it can never count the infinity of the variety of ways in which people end up in need of money in society.
Petersen: So before this welfare state that we currently have---the welfare state as it currently exists largely is a creation of the 20th century. But in the 19th century and early 20th century a lot of what you had was mutual aid societies and things like that. And I think a hard core maybe a Rothbardian libertarian who maybe still cares a lot about the poorest among us might say, why have a basic income? If we just had nothing there would still be the civil society and we could create something like a mutual aid society. Are there advantages of---is there reason to do this through the state, I guess is my question.
Lehto: As a very wide-going and deep-reaching utilitarian, for me it's all about checking what robustness criteria institutions might have, and what institutional arrangements we could come up with and seeing how they perform in the real world rather than in the realm of ideal theory. And we have some evidence of places where mutual aid societies worked and we have some evidence of places where forms of welfare state that are highly bureaucratic and oppressive and paternalistic have operated and both of those have several features that I think we can wish to want to get rid of.
So I think that if we look at societies where mutual aid societies were the sole means for people to survive I think they actually did a relatively good job in many cases but I think they failed to provide the sort of guarantee of security that I think a good society would wish to provide for people. That is, if we rely on the means of mutual aid societies you will get perhaps even a superior alternative to many forms of welfare state in the long run and I'm completely open to the idea that free markets can provide a very robust system of welfare. And actually that to me is one of the reasons why I consider myself a libertarian defender of a welfare state because I think that the libertarian part comes from actually understanding that markets are a good way of producing welfare and the opportunities available for markets and other forms of voluntary transfer, including mutual aid societies, are a way of providing a wide framework of security and services and other forms of protection.
But I think that they provide a patchwork which leaves a lot of people outside in a number of circumstances. And I think this fact that they have a lot of holes in the way into the system, they have a lot of uncertainty about guaranteed income and lot of uncertainty about who gets covered, who is seen as being worthy of being helped, who is seen as being worthy of being protected by a benevolent charity and so on, means that we need to have a system of making sure that people don't---perhaps out of no fault of their own---fall through the cracks of the free market system and the same goes for the welfare state. I think they actually are surprisingly similar the welfare state and the free market utopia, they both provide this patchwork framework where some people are protected, some people are not, there's a lot of uncertainty about who gets what, who gets protected, and who doesn't. And so actually in both systems, people fall through the cracks and this is exactly the reason why I think basic income guarantee can be a superior alternative to either of those.
But again we have to see what happens when we actually implement basic income, there could be a lot of unintended consequences. So we need to take those into account as well but at least on the side of theory, I think the idea of guaranteeing basic income, I think it's both desirable and practicable because we know how to do it technocratically and theoretically. I mean there's nothing so difficult with guaranteeing basic income via bank transfer to all citizens for example.
Petersen: So one virtue of the basic income guarantee is that it seems to be actually politically feasible within our current system and it has got some interest in recent years. We mentioned at the start of the episode that you were part of the effort to bring a basic income to Finland so could you tell me about the political situation there? I've heard that they're looking at bringing in a basic income guarantee.
Lehto: Yes, indeed. And here I'm being brutally pragmatic. Finland is not going to turn into any sort of libertarian utopia that I would wish for and certainly there are elements of paternalism there that are not going to go away. We still regulate the sale of alcohol in a very, I think, outrageous fashion for example and there are a lot of elements in the system that probably will keep us on the level of adult children for a long time. But as far as the welfare system is concerned, there is considerable consensus now that something like basic income would be a desirable reform. And this is seen by the majority of the population and by more than 50% of the M.P.'s in the parliament, basically from all parties with of course different proportions in different parties.
But yes indeed the center-right government is actually going with the basic income pilot experiment starting next year. It's I think a well-planned pilot. They have a lot of experts because we believe in experts in this country and in Finland the sort of reliance on experts is both good and bad in many ways. It always seems to suggest that there is a group of people who can define the perfect system but in this case I think they've done a pretty good job with planning this two-year pilot. We shall see what happens. It's certainly not ideal and the government is already bungling with some of its promises and how it is going to be organized.
But the basic premise for people who may not have an understanding or an idea in their mind of what this actually means, it means that basic income in the Finnish context would be the guarantee of something on the order of 500 to 600 to perhaps 800 Euros per month per person. And this would replace the various forms of unemployment benefits, sick leave benefits, student benefits and various other forms of benefits and Finland obviously has a lot of those already in place. And the complexity of the bureaucracy is such that even the experts who run it are surprisingly candid about their ignorance, about the complexities and mutual dependencies of the various benefit structures so that it's a maze that not even the experts can navigate, let alone regular ordinary people who are supposed to be the beneficiaries of the system. So a lot of people don't know how to apply for help, a lot of people don't know what benefits they're entitled to, and there's a long delay in getting the results of one's application for particular benefits---months, sometimes even the years. And a lot of people fall through the cracks in that fashion that I mentioned earlier.
And so I think we've come to the point almost by necessity, where this system is seen almost universally by all as in need of reform and basic income happens to be the form of this reform that is most universally seen as the one we should pursue even though of course there are still people who are very skeptical of it in many ways. But yes, indeed they're planning this experiment where they're giving something like 500 Euros to a few thousand people across Finland. It's a very small experiment, but there are people who will call for its expansion I'm sure in the years to come. That will be definitely a very interesting experiment to see how that goes.
Petersen: It seems like with the current system being so complex, it's almost like a part-time job just to collect benefits. You need to build expertise and you need to fill out the right forms and it takes a lot of your time and in many ways that makes it something that competes with the labor market for your time and your efforts and your human capital development. Seems like a basic income would be a good way to get people back into the labor market simply by virtue of freeing up their time to pursue something else. Do you see the political movement towards basic income making progress in other countries as well?
Lehto: So yes experiments are undergoing in a number of countries. In addition to these, Netherlands, Canada and U.S. experiments and the Finnish case of obviously which I'm most familiar with, there is a very interesting experiment going to start in a few years in East Africa organized by the charity Give Directly who are already advocates of this idea of giving cash transfers to people. They have been doing that for a number of years now with quite good results according to many independent researchers. They've been giving cash transfers directly to people and they've shown great results. So they are actually expanding this idea and organizing again a privately funded experiment that they planned around for ten years, I think, or at least a number of years in East Africa.
And this should be quite interesting to see how the basic income experiments in rich countries and poor countries compare and perhaps they can help both in different ways, because obviously countries where welfare states exist are quite different from places where they don't. So any help or any form of monetary transfer will help people in African countries proportionally more than they do in rich countries, but I think both situations and both contexts can certainly benefit from direct cash transfers and basic income.
Petersen: Give Directly is a charity that I support and I really like what they're doing. I especially like how they take such a quantitative approach. There are so many charities that just start with "wouldn't it be nice if people in this village had this thing?" And then they bring it to them and they don't really stop to say can we measure, were we cost effective in improving their lives? Did we do a good job? Could something else of equivalent cost have made them better off? Give Directly is doing a great thing by bringing a lot of this sort of quantitative approach to charitable giving and I'll have a link at the show notes page to Give Directly if you want to contribute, if any of the listeners want to contribute, I highly recommend it.
Lehto: Absolutely. For a little bit, just to say about the reasons why cash transfers are so great. By the way, I should say that there are perhaps a few charities that are even more helpful in certain contexts. For example, direct malaria helping efforts, efforts to eradicate diseases perhaps, have an even higher rate of efficiency but those are pretty much the only ones that are more effective than giving people cash. And the reason why giving people cash is very good is that first of all, they stimulate markets where they don't exist and where markets do exist they operate in a way that maximizes the preferences and satisfaction of the people concerned. They operate as a way of giving people welfare in the most efficient way possible.
And the theoretical foundations of these can be found for example in neoclassical economics, of course, where the superiority of cash transfers have been posited for example in the Chicago School since George Stigler and Milton Friedman and others. There's a wonderful paper by Brennan and Walsh on the desirability of cash transfers over in-kind transfers from a game theoretical Pareto perspective. So that's also quite interesting how the theory also matches the empirical research here.
And just again to go back to the very foundation of the welfare state. I think that's been the biggest mistake of the welfare states today that they fail to take into account how welfare truly fundamentally is the satisfaction of the desired ends and needs of the people themselves as they themselves see them. It shouldn't be the satisfaction of some criteria of goodness that the state bureaucrats measure and determine. It really should be ultimately up to the people themselves what they value, what they pursue, and what needs they see themselves as having and thus giving money to them is the best way to make sure that they actually get to satisfy those preferences which they have rather than those preferences which some bureaucrats think that they should have.
Petersen: If I may ask one final question. Some supporters of the basic income guarantee have suggested that we could do it as a swap. We get rid of our current costly welfare system and bring in the basic income guarantee and often you'll hear the suggestion that this could be revenue neutral. Is that a realistic possibility?
Lehto: It is a realistic possibility in cases where quite extensive welfare states already exist. And obviously it depends on the level of basic income and I'm actually in favor of starting low where that is the most politically feasible option. But I'm also quite a quite supportive of the idea of starting high where that is politically feasible. So in countries like the welfare state in Canada and many other places. Starting from the level of where the current welfare state benefits are it is compatible with the goal of making it neutral as far as the effect on state budget is concerned. Although I think that it will be very hard to make it completely neutral in that regard. I think it will by necessity always cost something.
But what it will cost is heavily overblown in many estimations because many people simply do not understand how to calculate the costs and they simply add up some figures of everybody gets this amount of money and multiply that by the number of people and voila you get the proposed cost of this program. But that's obviously nonsense that they don't understand what they're talking about. And they really should have a look at the actual models because in all models what happens is you reform the tax system at the same time which means that for most people, middle-class and upper-class people---or middle income and upper-income people, to be more politically correct---the income that they get from basic income actually is a zero sum addition because actually, they ended up paying their basic income back in the form of taxes that I've been raised to match accordingly the need for basic income funding. So, even if there is no criteria that you don't give basic income to people above a certain range of income, nonetheless those people in the upper brackets will end up paying back their basic income due to the taxes that have been raised. But the taxes that are raised do not have to impose unbearable burdens on those people either, because again for most people it is just a nominal transfer of funds and it's withdrawn from their bank accounts at the same time.
Petersen: So are there websites, books? What can you recommend to people interested in this topic? What should they read?
Lehto: Well I think for those who are philosophically minded, I certainly recommend reading the classics of the libertarian welfare state stuff. Things like Friedman's Capitalism and Freedom where the negative income tax fee is featured. Friedrich Hayek's Constitution of Liberty is a great book and it also features a defense of guaranteed minimum income. And more recently John Tomasi's Free Market Fairness, and I would recommend people to read the blog Bleeding Heart Libertarians they have been advocating for basic income but also debating it. And also proposing this similar thing that I'm doing which is trying to combine Rawls' and Nozick's intuitions into something like a new coherent whole.
And just follow the news, read up on the models, follow up on what the governments and many of these countries---Finland, Netherlands, Canada---are doing. And go to basic income networks website. Just Google basic income earth network. B.I.E.N it's called---Basic Income Earth Network---and you will find more about basic income.
Petersen: My guest today has been Otto Lehto. Otto thanks for being part of Economics Detective Radio.
Lehto: My pleasure. It's been fun.