The Success of the Embedded State

James Robinson interviewed by Tim Philips

James Robinson  00:01

70% of people working, providing public goods, judicial services, or whatever, were unpaid.

Tim Phillips  00:09

How does a government function on almost no budget? My guest today has examined the historical role of unpaid civil servants with some surprising conclusions, so today on VoxTalks Economics, the success of the embedded state. Welcome to VoxTalks Economics from the Center for Economic Policy Research. I’m Tim Phillips. Modern governments spend roughly 45% of GDP, and among other things, that money buys a professional civil service to do the bureaucratic tasks of government. But this is a recent phenomenon, how did that bureaucracy function when tax revenues were tiny? James Robinson of the University of Chicago is one of the authors of a new paper that focuses on what he calls the “embedded state”, and he joins me now to explain what that is, and how it evolved. James, welcome back to Vox Talks Economics.

James Robinson  01:25

Thank you very much, I’m delighted to be here.

Tim Phillips  01:29

James, you’re telling the story of England’s local government bureaucracy, and we’re going all the way back to the start of the Industrial Revolution. Around that time, what sort of functions would these civil servants be providing?

James Robinson  01:43

I think it’s important to understand that their civil servants are different levels, you know. There’s national civil servants, people working for the Treasury in London, and then there’s people working for the counties, there’s people working for rural parishes, and there’s people working in the city, but if you look at the city government, you know, the so-called boroughs, for which we have very good data, you can see people are doing all sorts of things, they’re keeping a jail, they’re policemen, they’re looking after the harbour, they’re building bridges and roads, judicial services are very important, so they’re holding court, they’re settling disputes, raising taxation in some cases, keeping accounts, lots of different local activities, providing local public goods.

Tim Phillips  02:26

This sounds very modern, but the government didn’t have huge tax revenues at that time, certainly nothing like it would have today. So, who is getting recruited to staff this civil service that’s doing all these tasks?

James Robinson  02:39

Yep, so that’s what’s so fascinating about it. The dominant narrative about British economic history comes from John Brewer’s marvellous book, The Sinews of Power, about the emergence of this kind of fiscal state after the Glorious Revolution in 1688 and he talks about raising taxes and paying incentives. And yeah, that’s all true for the 20,000 people who were working in London, but it turns out in the rest of the country — where, as you say, there was very few fiscal resources — most people basically work for free. So, all people who worked in parishes were unpaid, and in our sample of boroughs, 70% of people working, providing public goods, judicial services, or whatever, were unpaid. The question is: How on earth do you get people to do things when you don’t pay them? You have to exploit the nature of the society you’re living in. You can get prestige by being an alderman, or being the Lord Mayor, or being the Chamberlain. Perhaps it’s a stepping stone to becoming a member of parliament or getting a job where there is money and even more prestige and status. When you look at these jobs, and you look at who’s doing them, elites do jobs in which they can get prestige and local status, and that’s enough to get them to do it, or it’s a stepping stone to becoming a member of parliament, or getting some other job, which is paid. But you can’t acquire status looking after the jail, so you have to pay the jailer.

Intermission  04:20

The warning that snitches get stitches is heard around the world. It’s easy to imagine why groups adopt pro-social norms like sharing or volunteering, but what sustains an anti-social norm like this one? This is what we talked about the last time James spoke to us in October 2024. Find the episode called ‘James Robinson on Antisocial Norms’ when you follow us wherever you get your podcasts.

Tim Phillips  04:58

So in this story we’re going to tell, the key document dates from 1835, so almost 200 years ago. What was that document? What’s in that document?

James Robinson  05:09

We’ve known about this phenomenon for quite some time. There’s a very brilliant British historian and political philosopher called Mark Goldie, who wrote an article about this 20 years ago, or something, even longer than that, would, so, which we stumbled over, and we thought, gosh, this is exciting, you know, I was using it for teaching, but how do you study it systematically? And then we discovered this 1835 parliamentary report. So, basically, after the 1832 reform act — when Britain became kind of much more democratic in a modern sense — parliament started trying to investigate lots of things, and it wanted to understand. It didn’t really know what was going on with these borough governments, so it hired a bunch of lawyers and sent them out all over the country to these, about 300 of the largest boroughs, largest cities, and systematically collected information on who was working for the government. What did they do? How well did they do it? That’s what’s so exciting. How much did they get paid? So, in this three and a half thousand-page report, one of these parliamentary papers, that’s the factual basis for what we’re doing. And it was really because parliament itself didn’t know what was going on. It’s stated very explicitly, because these borough governments developed historically in the Middle Ages. They might have had charters from the king, and there’s a lot of idiosyncratic variation in the way these borough governments were organised.

Tim Phillips  06:37

In doing this, you want to estimate how effective the paid officers and the unpaid officers were. Compare the two, but to do that you need to have some measure of the quality of the service they’re providing. So, from 200 years ago, how do you measure the quality of courts or how good the policing is?

James Robinson  06:56

So, the reports do that. The reports go through, systematically, everybody who was working for the borough, what their tasks were, whether they were paid or not, if they were paid, how much they got paid. But then there’s a lot of discussion about how well everything was done. So, the parliamentary commissioners graded: does there exist a jail? If there exists, is it in satisfactory commission? Quite satisfactory, very satisfactory. There’s actually systematic evaluation of the provision of these public goods, which the borough is tasked to provide, you know. So, they say whether they provided at all, but then if it’s provided, then they actually evaluate the quality with which it’s provided.

Tim Phillips  07:37

You’re trying to allocate that quality to particular individuals?

James Robinson  07:43

What’s fabulous about the report is it tells you who’s responsible. You know, who’s responsible for dredging the harbour or who’s responsible for the streetlights being lit. So, we can see who’s responsible for street lighting, and were the streetlights at all, and if they were, were they satisfactory? So, we can exactly compare those two things,

Tim Phillips  08:02

But if we did this in 2026, then it would be very hard to do, because everything’s delivered by teams and it’s hard to allocate quality to a member of a team. Did you run into that problem?

James Robinson  08:13

Absolutely. Yeah, so multiple people could be responsible for the same thing. In the paper, we use a kind of econometric model of teams, people are jointly working to produce some particular outcome. And then there’s strategies for trying to evaluate what an individual’s contribution was to that team. You know, what’s fascinating is there’s a lot of variation here in terms of who’s on what team, and some boroughs, the alderman is responsible for one task. In another borough, they’re responsible for a different task, and so there’s a lot of variation here that we can exploit to try to get at that issue.

Tim Phillips  08:50

Okay, let’s go for the big result first. What was the performance gap between people who were paid to do the job and people who weren’t paid to do the job?

James Robinson  09:00

We find that people who are unpaid are actually much more efficient than people who were paid. Before you draw the wrong policy conclusion from that, and sort of say, well, we should stop. Let’s not…

Tim Phillips  09:11

Let’s not pay anyone, then.

James Robinson  09:12

Of course, the reason is that very different sort of person was doing unpaid jobs, you know. Elites were doing that, and they were motivated. They were incentivized differently. They were not incentivized by getting paid; they were incentivized by prestige, by things like that. You know, it’s very interesting, actually. In the report, it ranks British people — they love this sort of thing — different tasks in terms of sort of prestige. So, unpaid people are more productive, but I think the way you have to think about that is that’s really because a very different sort of person was doing unpaid tasks and they were incentivized differently, but it shows that those ways of incentivizing people was actually very effective.

Tim Phillips  09:54

And is this a stepping stone in their career? So, when they actually get the job, that isn’t the end of the story for them. They actually have to carry on doing the job well, so that they can move on up the ladder.

James Robinson  10:04

Absolutely, you know, so we can merge this data with the history of parliament, for example, and look at who, you know, do people show up as members of parliament subsequently, and there’s qualitative historical information we discuss in the paper, also. But yeah, we can look directly at that.

Tim Phillips  10:19

I would imagine in the work of trying to find this relationship, you’d run into the problem that some places would have a lot of unpaid positions but would also perform well. A bureaucracy would perform well for different sorts of reasons. Can you be sure that this is a causal relationship we’re finding here?

James Robinson  10:40

The way we approach that is to say, if you look at the history of these boroughs, and there’s a lot of variation in the fiscal resources that they have, that variation is historically determined. It’s determined by their medieval charters, which allowed them to raise particular taxes. Often, they inherit land, you know, people die and they bequeath land to the borough, and the borough rents it out, so that generates income. We try to narrow down to these kinds of historical sources of variation in your fiscal capacity and use that as a source of variation in whether you could afford to pay people. I mean, there’s different sorts of variation here, you could imagine, but it seems like if you read the historical work, for example, the nine-volume history of British local government. I’m sure you’ve read it. I’m sure I’m sure everyone listening has read all nine volumes by Beatrice and Sidney Webb. They actually talk a lot about these kinds of historical constraints. Now that’s a source of variation in your ability to pay people, so we use that to try to get at more causal relations, you could say.

Tim Phillips  11:48

I guess, if we could go back in time and ask these English gentlemen who were working as bureaucrats what motivated them to do it, they might say, well, it’s a selfless desire to provide good government to the regions of England. You’re saying that actually, really, there’s a.. it’s the opportunity for promotion or self-promotion here. Does that make a difference? Does it matter?

James Robinson  12:12

I don’t know. I mean, you know, this is very hotly contested by historians, you know. If you read the work of John Beckett, who was a very distinguished British historian of the aristocracy, he really likes this idea that the British aristocracy were a service elite. There was a social obligation to undertake these tasks. I think, like, nowadays it’s hard to recapture that moment, you know, like for example, we look at gentry, like, who was a gentleman, and we show that gentlemen are much more likely to occupy unpaid positions, but like in contemporary British society, it’s hard to capture that notion of a gentleman. No, it doesn’t really exist anymore, like it’s disappeared. You have these more specific titles, like a knight or whatever. Gentleman means nothing in contemporary British society, really, does it, I mean?

Tim Phillips  13:01

I don’t think many people would say there are many of them in politics now. That’s for sure.

James Robinson  13:05

Well, I wouldn’t be able to comment on that.

Tim Phillips  13:08

Absolutely. Your unpaid elite, though, you also did find that it was more likely to embezzle. There was more corruption. It practiced nepotism. Could we sum this up as a more corrupt system that somehow worked?

James Robinson  13:25

Yeah, I mean, that’s certainly one of the motivations for the parliamentary inquiry was this sense that there was like corruption in the boroughs. The 1832 reform act was the measure that eliminated the so-called rotten boroughs, for example. So, I think there was a lot of concern about those issues, and so the parliamentary commissioners were tasked to record information about corruption and nepotism. You know, was the mayor hiring their relatives and things like that. So, we can look at that too, and that does seem to be going on. So, we found that realistic, you have to give these people a lot of slack and you’re not paying them, they’re going to indulge a bit. What’s the net effect of that? Well, you know, we still find that the public goods get provided, so you could take the perspective that that was an inevitable side effect of running the local government this way.

Tim Phillips  14:14

So, James, what do we learn from this, and particularly, what do we learn from this about the sorts of things that you have studied and written about in your career? Now, you’ve long, very famously argued that formal institutions, well, they’re the sort of bedrock of development. Is this paper undermining this? Was 1835 England an exception to the rule?

James Robinson  14:43

No, I don’t think so. Actually, I mean, I think what’s fascinating is that this is probably what goes on everywhere. One thing that Leander Heldring and I have also been studying is the organization of the state in Rwanda. And most people who work for the government in Rwanda are not paid either. The International Monetary Fund says if you’re going to have a proper government, you need to raise 15% of national income in taxes. The Rwandans have never done that, but they have this amazing state capacity and ability to implement policy. And how? Well, because people aren’t paid, you know. So, so, and they collaborate with the state, and they’re integrated into the state, and I think that’s the story sort of everywhere. And I think that has very important policy implications, because everyone is so fixated on taxation and raising taxes, and okay, fine, you know, but actually that’s not the way many states achieve things. But they achieve it in different ways. And this is one of the things which I find so fascinating, meaning the way Rwanda does that, informally, is very different from the way Britain did it in the 1830s, or the way South Korea did it in the 1970s. And that’s why this word kind of embedded is important, like the state is embedded in society, but the sociology of different societies is very different, and that means that embedding is going to look very different, and it means you have to understand something about the sociology of the society to understand how the state really works, and I find that very exciting intellectually. Does it undermine my previous work? No, of course not. I mean, formal institutions are just much easier to measure and compare, and I think in social science, to make progress, you have to focus on simple stories and things you can measure and present data on. And there’s always layers of things that you don’t understand. A lot of the work that we’ve done, or other people have done, also is we focus on things like the capacity of the state, and some things you can measure, you can measure bureaucratic principles, whether people get hired on some meritocratic criteria, is their examination. How do people get promoted? Like, what does the wage structure look like? Are there, you know, so you can measure things like that, but it turns out there’s many other things going on, which make up something called state capacity, and I think that this is what we’re trying to study. I think you know, you peel back the layers of the onion, and often you start at the place which is easiest to study, and where the data is better, and then you just keep going.

Tim Phillips  17:11

Can we draw out lessons from this piece of work into how good governance might work today? As you said, it’s very different in Rwanda, but it has some things in common. Are there rules? Are there good practices that we could think about?

James Robinson  17:27

Yeah, I mean, that’s a great question. I’m not sure I have a great answer. I think the message here would be, of course, many groups of people organize to provide public goods collectively, and you know, Eleanor Ostrom’s work, for example, for which she got the Nobel Prize, was all about that. But I think people think of that as being like something separate from the state, you know, it’s something kind of antagonistic to the state, almost. I think what we’re showing is there’s a real productive interface between the state and things like that. And I don’t have a best practices, you know, but, but it seems like I would say, you know, if you were working for the World Bank, or you’re working for the British Development Agency — if that still exists, and they haven’t shut it down — you’d start by saying, okay, I want to improve public good provision. I want to get things working in Somaliland, what do I do? And instead of just saying, “Okay, I need to raise more taxes, I need to figure out how to implement a value-added tax”. Start thinking about society and who’s providing public goods, and how are they doing it, and how are they connected to the state? And Somaliland is an interesting example, by the way, of course, because Somaliland has this very elaborate clan structure. The Senate of the country, the Gutierre represents the clans. The sociology is like there represented in the Senate of the country. Just start thinking about things like that, and start thinking about what kind of potential for providing public goods does that create?

Tim Phillips  19:09

It creates huge potential, doesn’t it? James, thank you very much for talking about this.

James Robinson  19:13

My pleasure.

Tim Phillips  19:23

The paper is called ‘The Success of the Embedded State in England’. Authors Leander Heldring, Davis Kedrosky, James Robinson, Matthias Wiegand. It is Discussion Paper 21460 at CEPR. James, it sounds like this is work in progress. You and Leander are going to be delivering some more research on this line.

James Robinson  19:46

Absolutely, yeah. We’re very excited about it, both intellectually and from a practical point of view.

Outro  19:55

VoxTalks Economics is a talk normal production. The assistant producer. Is Megan Bieber, and our editor is Andrei Zagarion. Next week on VoxTalks Economics, helping the over 50s find a job.