This week we’re looking at the role of information in state-building. States need to be able to act. For that, they often need information on their citizens. But how do they get that information, what are the implications for how well the state functions and what do citizens get in return?
One of the most basic questions regarding any state is 'can it act?' Does it have the capacity, that is, to uphold the rule of law and to deliver security and public services?
For a state has the capacity to act it needs information on its citizens. You can’t tax someone or assess their eligibility for services if you don’t know who or where they are.
But states may be unable to require its citizens to provide information – it may have to rely on their wanting to do so. And that has potentially profound implications for how equitable state activities are – and therefore ultimately how the state develops and builds its legitimacy.
We are joined this week by Dr Jeremy Bowles, Lecturer in Comparative Politics, expert in the political economy of development and the interaction of state-building processes with distributive politics.
Mentioned in this episode;
[00:00:00] Alan Renwick: Hello. This is UCL Uncovering Politics. And this week we're looking at the role of information in state building.
States need to be able to act. For that they often need information on their citizens. But how do they get that information, and what are the implications for how well the state functions?
Hello. My name is Alan Renwick. And welcome to UCL Uncovering Politics – the podcast of the School of Public Policy and Department of Political Science at University College London.
One of the most basic questions regarding any state is: can it act? Does it have the capacity, that is, to uphold the rule of law and to deliver security and public services?
Whether a state has the capacity to act depends on many factors. One of the most fundamental is whether it has information on its citizens. You can't tax someone or assess their eligibility for services if you don't know who or where they are.
But a weak state may be unable to require its citizens to provide information. It may have to rely on their wanting to do so. And that has potentially profound implications for how equitable state activities are, and therefore, ultimately, how the state develops and builds its legitimacy.
Well these fascinating and incredibly important issues are explored in a new article by one of my newest colleagues here in the UCL Department of Political Science. He is Dr Jeremy Bowles, Lecturer in Comparative Politics.
And I'm delighted that Jeremy joins me now in the studio. Welcome, Jeremy, to UCL Uncovering Politics.
And so we've got a lot to unpack in our conversation today, even just from that brief introduction to your work. I think we can see that there's a lot of lots to get our teeth into here.
Let's start then by laying some foundations. What exactly do we mean by state capacity, and why does it matter?
[00:01:45] Jeremy Bowles: I think that you did a great job describing it yourself, but I'll have a go.
I guess there are often as many definitions of state capacity as there are authors who have written about state capacity. But I think that, like, an easy one to start with is to think of capacity as just the ability of the state to do the stuff it wants to do – to implement policies, whether these policies are around delivering services or extracting taxation or coercion or welfare provision.
I think there's been a gradual evolution in terms of how people conceive of state capacity, where older conceptions maybe focus more on the outcomes of capacity – kind of governance outcomes, you know, for example, literacy.
But more modern conceptions, I say, think more about what are the fundamental resources that states need in order to administer any kind of policy. So rather than thinking of capacity as constituting taxation, maybe capacity constitutes the resources a state needs to collect taxes from people.
And so if you think about state capacity as having these resources required to administer a broad array of different policies, then naturally you start to think about like what are these resources you need to administer these kind of policies.
And as you alluded to, a quite important one that lots of – a very distinguished lineage of – scholars have written about is information – that you basically need to know stuff about your territory, about your citizenry, in order to tax them, to deliver services to them, to administer welfare to them as well.
Other ones that people write about are, for example, fiscal resources as a form of capacity, or the bureaucracy as a form of capacity as well.
It's obviously a multifaceted concept, but at least in this paper – and a lot of my work – I focus on these informational resources as being kind of an important aspect of capacity.
[00:03:32] Alan Renwick: And how kind of difficult is information as an element of state capacity?
I mean I guess, me, sitting here in London, being used to how things work in the UK, I kind of take for granted quite a lot of information that is just very widely available. But I suppose in a country where we're thinking about building this state capacity often that's not the case.
[00:03:51] Jeremy Bowles: Sure.
So, or historically that's not the case as well. So actually even in the UK, a lot of systems we think of as providing information to the government are actually fairly recent. So things like civil registration, for example, passports – a lot of these are 19th or 20th century creations.
By contrast, in a lot of developing countries today, there's far less information about us available just to our governments.
One reason for this is just that citizens interact less frequently with their governments. Another reason for this is that there are fewer formal technologies for governments to learn about citizens.
So some of my work, which we'll talk about a bit, thinks about civil registration. So foundational documentation, things like birth certificates. And even administering birth certificates is very expensive for a government because you have to be continually operating a bureaucracy capable of registering people as they're born.
Other examples you might think of are the census, which historically are incredibly expensive, often quite contested, political acts. And still today in developing countries, it's extraordinarily hard and expensive to run a census, which only happens every 10 years typically, partially because it's so expensive to do so.
And so while we might think of ourselves, at least in the UK, as being, you know, the government prospectively knows quite a lot about us, either historically today even in the UK, or across much of the developing world, it's easy to find examples of where governments don't know very much about their citizens.
[00:05:16] Alan Renwick: So we know that state capacity matters. We know that information is an important aspect of this.
What's the kind of problem in connection with this that was driving your research? How did you end up getting interested in this area?
[00:05:29] Jeremy Bowles: Sure. So, I guess two directions.
On the more academic side, there's a lot of this classic, really great work in political science thinking about how information – or legibility, you want to say, to follow Scott's famous term – how information affects governance, the downstream consequences of states knowing more about their citizens, and how important this is for state consolidation and political development and so on.
There's much less thinking about where it comes from, about like how do states actually develop this capacity. And a lot of the cases that are studied are basically in the context of quite strong, fairly coercive states which are able to go out and identify people, register people, enumerate them, and so on.
There's much less studying of context where states are weaker, basically – where we're often in developing country settings.
And so the academic motivation was to think more about like why is there so much variation in these informational resources if we think from this classic literature that they matter for so many different governance outcomes. You know, what can you do if you don't know anything about people?
On the more real world case, this was motivated by some summers I spent in Liberia during graduate school, where I became interested in the efforts of that government – as many other developing country governments are doing today – to try and roll out new, quite technologically sophisticated schemes to identify citizens.
Because in a lot of the world's poorest countries today are where you'll find the most sophisticated schemes to try and identify citizens. Whereas, you know, people in the UK would protest if there was like a national ID scheme, there are basically no countries in Africa which don't have a national ID card scheme nowadays.
And I became interested in, like, why are these schemes so popular, but often sort of fail or find it very hard to reach high levels of enrolment among the population.
And spending some time in Liberia through graduate school got me interested in this from the policy aspect of talking to the relevant government ministries there, and I ultimately thought that I'd take a slightly more historical lens on this to think about prior efforts by governments across different parts of sub Saharan Africa to register citizens to think about some of the dynamics there.
[00:07:39] Alan Renwick: So you actually look in this article that we're focusing on today on at Tanzania, and particularly Tanzania in the post independence period, so the post colonial period, starting in the 1960s.
Why was it that you chose to focus there?
[00:07:54] Jeremy Bowles: Yep. So a few reasons.
I guess the first is that it's a little bit forgotten today, I think – and at least in the literature – but the post independence decades were characterised by a real sort of developmentalist push in a lot of post independence African states, where often these states inherited extremely little from antecedent prior colonial regimes, either in terms of bureaucratic infrastructure or information or fiscal resources. They inherited basically very weak states. And a lot of these states then exerted real meaningful effort to try and kind of build under the constraints they've inherited.
And so just as today there are all these countries which are rolling out these quite sophisticated schemes to try and identify citizens, actually there was also this window in the 60s and 70s where a load of different African countries were doing the same thing.
And it's a little bit forgotten today because the donors involved were a bit different and obviously, in general, people don't study that era as much today obviously as the modern era.
But there are lots of these efforts in different countries back then.
And so I thought that an interesting way to speak to potentially some of the trade offs that governments face today is to look back in time and to think through some of the causes and consequences of these schemes in the early post independence period.
So that's kind of why historical because, you know, it's easier to trace out the impact of these things when you can actually have years afterwards to observe versus stuff which is happening right now today.
Why Tanzania? Tanzania is like an interesting – like a super fascinating – case for a lot of reasons post independence, politically, but also in terms of the sort of developmentalist impetus of the post colonial regime in Tanzania, which really tried quite hard to implement these schemes.
However, if you look at Tanzania today, that might be slightly outdated. But roughly it's like 10 to 15 percent of the population that has a birth certificate in Tanzania.
[00:09:51] Alan Renwick: Wow.
[00:09:51] Jeremy Bowles: It's incredibly low. Any type of formal ID is also extremely low.
And so I thought it was interesting to say, like, here's this country which in the post independence era was like held up as this ambitious modernising state, at least in the 1960s, but today, actually, rates of enrolment in these kind of informational schemes are extremely low.
And so that motivated me to read more into the case.
[00:10:17] Alan Renwick: And so the core idea here is that actually it's quite difficult for states to get citizens to provide this sort of basic information on themselves.
So I guess one aspect of that is that a relatively weak state may not be able to require and enforce rules saying you must register births and that kind of thing.
And presumably also in a context where state legitimacy may be fairly low, then people don't kind of spontaneously want to do this because it's good for the community as a whole. There isn't that sense of attachment to the state. So therefore you're relying on people thinking that it's in their own interest to provide information to the state.
And that's kind of where you start in looking at this research – that we need to rely on people's interests. And that then sets up difficult incentive structures that states need to engage with.
Do you want to just tell us what your starting thoughts are there?
[00:11:18] Jeremy Bowles: Yeah, I mean, well, well summarised – much more clearly than myself.
But yeah I think you're exactly right. The sort of the starting point here is to think that if we're in a state where it can coercively extract information from citizens, then that's one case where, you know, a state where you can just send people out to register or enumerate or whatever you want to do to people. Then that's sort of like one case.
However, that's not relevant to a lot of particularly relatively poor countries. And if we're in these sort of contexts where coercion is sort of off the table, as it were, then really you need to rely on it being in citizens’ own instrumental interest to comply with state demands for information or for other kind of policies as well.
And so the starting point here is to think about what are these instrumental incentives going to look like to enrol with the state when it's like, I don't know, register your births, get an ID card, this sort of thing. And what does that trade off look like?
There's a very simple, stylised way to think about it, but the main benefit of these kind of things for a citizen is like, what public services, basically, do I get if I comply – do I get access to – which might be something to do with education, or healthcare, or social services.
And secondly, what are the costs?
And the costs that I focus on here, which are particularly salient in this post independence period, were to think about taxation, basically – that if I like give information about myself to the state, maybe I get access to something, but also potentially, you know, I'm opening myself up down the road to being taxed in these systems which inherited quite extractive systems of taxation.
[00:12:52] Alan Renwick: And therefore the opening sort of expectations that you have are around what the state will do in order to create incentives for people to register and so on, and what the implications of those incentive structures are.
[00:13:09] Jeremy Bowles: Yeah.
Basically, like if we think about this instrumental decision facing citizens, plus having a context of a state, which, you know, is at least kind of like intermediate capacity, then how do we... Sort of what patterns might arise from these two contextual features.
And what I focus on in the paper is, okay, so how do states actually generate incentives to comply with demands, given that citizens might face some downside cost or risk of registration. And that's that they gatekeep – that they withhold access to stuff in the absence of documentation, for example, to say like, Oh, you can't claim a social cash transfer or something, unless you've got a document.
But then the question is, well, what kinds of services can they actually restrict access to? And again, in a very strong state, it's very hard for me – and for us in the UK – to do much to do with the government unless we prove who we are at every kind of interaction.
But in a slightly weaker state environment, that's not feasible – it's not, kind of, enforceable to always demand documentation for accessing state resources. However, it's relatively easier to restrict access to stuff on the basis of documentation to things which are maybe cheaper for the state to gatekeep, which might be things like, I don't know, access to government jobs, more formal sector social security or aspects of education where there are, you know, fewer facilities to manage as it were.
But secondly, where if you gatekeep access to these resources that you are basically attracting the compliance of the people that you sort of want to be able to know more about because potentially you can tax them as well. And so that basically induces some kind of structural incentives to say: not only do I want to restrict access to universities on the basis of documentation, because it's easier for me to manage 20 universities versus a thousand schools, but also, if I do that, that I'm learning more about the people who are going to go to university, who are going to be the people who are more sort of fiscally valuable for the state to know about.
And so the first side here, then, what this kind of implies, is that there are quite stratified incentives – economically stratified incentives – facing citizens to decide whether to basically comply with state demands for information, which are almost embedding a kind of economic inequality.
And what the empirics of this paper are trying to do is to say that like: how can we just try and think about these stratified – economically stratified – incentives, and not all of the other slightly more mechanical ones that are also going to go in the same direction.
So, for example, like, richer people live in more urban areas which are closer to schools and identity registries, say. For sure that's true. However that's a less sort of political story, as it were.
And so what I'm trying to do in the paper is to say that, like, okay, let's try and, like, ignore that other stuff about rurality and so on and about ease of registering yourself. Let's just try and isolate the differential incentives that relatively wealthier people might face to register with the state with a view to the fact that these incentives are set by a government that's kind of weakly constrained in terms of what it can actually restrict access to.
[00:16:21] Alan Renwick: Great, thank you.
And we'll get into the data and what you find in just a moment. But just before we dive into the detail there maybe just sum up where we've got to so far.
So, essentially, you're arguing that states need information on their citizens. In order to get that information, they need to give their citizens an incentive to provide it. That incentive will often take the form of restricting access to things that citizens might want: public services only to people who have certain documentation, i.e. have provided certain information to the state.
And, I guess just before we go into the data, it would be useful to understand why this matters.
So I think you're also suggesting that this kind of causal chain that we've just set out there actually has some potentially rather worrying normative implications for how the state functions.
[00:17:16] Jeremy Bowles: Right. Yeah. So it's not the... So exactly as you're alluding to, it's not just that these kinds of schemes and this kind of informational capacity embeds – reflects – economic inequality, but it also potentially entrenches it, with the idea that like once you've actually complied with these things, people only comply with them because it gives them access to particularly useful things, maybe for richer versus poorer people. And that therefore rather than just like reflecting inequality over time these things might actually entrench it as, you know, maybe as the state expands what you can get in order to after having registered and so on, that it sort of amplifies pre existing inequalities.
And so when we look at variation in terms of like how many people are legible to the state, it's not just about... Firstly, it's not just about whether they live in areas where they face lower direct costs of identifying themselves, but rather that it reflects this sort of this trade off which is set by the government. But secondly that that trade off embeds and potentially enhances economic inequality as these kind of informational tools become more useful for accessing particular things.
[00:18:27] Alan Renwick: Good. Let's focus then on the research that you've actually done and reported in this paper.
So, as you were suggesting earlier, there are all sorts of factors that might influence people's access to state services you were mentioning there, whether they live in urban and rural areas, for example. So all sorts of things. And you're trying to isolate the impact of this information issue that you've identified.
Do you want to just quickly describe for us how you go about doing it?
[00:18:57] Jeremy Bowles: Yeah. So in short, like what I'm using, at least in the paper here, is the fact that Tanzania in the early post independence period. So basically its independence was 1961, and then around ’65, ’66, there are quite a lot of initiatives that the government implements to try and get information about people because they realise that this is a big problem.
One of the kind of interesting contextual triggers here is they run these elections in 1965, and nobody wants to register to vote because they're too afraid they’re going to get taxed if they register to vote, because there's such a strong association between any act of registering with the state and getting taxed because of how colonial taxation systems worked in Tanzania.
And so around, sort of, ’65, ’66, the government starts to be quite concerned about, like, basically not knowing very much about who people are – being able to identify them. And there are reports in ’63 and ’64 around how it's hard to administer taxes when, for example, some taxes only kicked in if you were above 18 or below 18. Then they didn't know – there was no way to tell, basically.
And so there was an increasing recognition that some of the information was imposing a constraint on governance.
And so what the government does in ’66 in a bunch of relatively more rural areas is that – sorry, more urban areas – is that they implement a set of these legal reforms – kind of weakly enforced legal reforms – to make it a nominal legal obligation to register births, to say that you now are going to be fined some amount of money if you don't register the birth of your child within a certain amount of time, where beforehand nothing existed.
And so at a high level, the empirical strategy is to think about like, okay, let's look at what happens then in these areas where it becomes a nominal legal requirement to register births and therefore give birth certificates – issue birth certificates. Let's look at what happens in these areas compared to in similar-ish areas that don't have this reform implemented, before versus after, to try and get some sort of empirical traction in being able to say like: Okay, first of all, let's think about what are the impacts of registering – of like, what do people observe later in life who are registered somewhat, you know, plausibly randomly at birth because of these reforms. Like what do they get access to – more versus less?
And secondly, now let's look back at the moment of the reform and think about like what determined, across these different areas, whether more versus less people actually complied with the government's demands to register people – to register births.
[00:21:51] Alan Renwick: And what do you find?
[00:21:52] Jeremy Bowles: So there are two main areas of the results that I focus on in the paper.
The first is to think about what are the consequences of registering with the state. And the second to think about what are some of the determinants of who complied with these reforms in light of some of the considerations we were just talking about.
And so in terms of thinking about what's the impact to you, think of this as an experiment where, you know, what's your life like if you're given a birth certificate versus not, I look at a couple of different sectors, including: how do your educational outcomes look different; how does your access to social welfare, social security, look different; and also, how does your exposure to taxation look.
And so on the first of these, the interesting distinction is between more universal primary education where there's basically no effect. People who are registered in this setting have no more or less access to attainment of primary education as observed later in life than those who aren't registered. However, they do have much more access to secondary and post secondary education, which in this setting is very much associated with the economic elite. Much richer families, for example, are those who are going to secondary schools, which is quite a small portion of the population during this period in the country.
[00:23:10] Alan Renwick: And just to clarify, you're saying that there's an effect of whether they're registered or not that goes beyond the effect of their being richer.
[00:23:19] Jeremy Bowles: Exactly. So the idea is that you can sort of purge the estimates of all of that other stuff, just to think of this almost like kind of quasi experiment to say that if you've given documentation and can prove who you are to the stage later in life, there are discernible increases in access to higher levels of education.
And the same basic pattern of results applies when I also look at social welfare and social security, where there aren't particularly strong effects on these more universal schemes to do with health insurance, but there are stronger effects on things to do with social security associated with government jobs and state employment, which again is very much associated with the economic elite.
And so these results are broadly consistent with some of these ideas around the fact that these informational technologies don't necessarily have this universalising access – on getting access – to more broad based public goods, like primary education, but actually they're more useful, at least in this setting, seemingly, for accessing more regressive, more narrow based public resources.
The other set of consequences I look at are to do with exposure to taxation, where I also find that basically people who are as if randomly registered, or are identifiable to the state, also pay more taxes. And this is particularly through more centralised modes of taxation, rather than more local modes, which tend to rely on sort of the central government having more information about you.
Part of this might come through having more formal sector employment, potentially. Part of it might be through other channels of direct taxation during this period which were administered by the government.
But basically, these people aren't getting more access to universal public goods or broad based public goods. They’re getting more access to narrow based public goods.
In exchange for this, potentially, they're also paying more taxes to the state, as observed later in life in the data.
And so this is to think through some of the slightly, the somewhat, the incidence of the benefits of registration.
And then the second part of the analysis is to think about, okay, what determined whether people actually complied with these reforms, these weakly enforced reforms, in the 1960s to begin with.
And for that, I’m basically looking at spatial variation in taxation, the threshhold of extraction, and also the presence of these different public goods around the time when these people were born, in the mid-60s, where I find consistent evidence to suggest that people were concerned about the threat of taxation – that areas where there was more taxes being collected, or where basically tax, you know, the tax schedule was particularly progressive, where the rich faced much higher tax rates than the poorer people. These are areas where fewer people complied with these reforms to identify themselves to the state, consistent with basically particularly richer people being the most concerned about identifying births, basically registering births to the state, because of potential downstream concerns around extraction.
However, areas where there are more particularly higher levels of schools, so secondary or post secondary education, these are also areas where people are more likely to comply. Which is to say that, like, provide some evidence that, yes, that first part of the results are to say we have these sort of stratified consequences of registration.
The second part is to say that actually there is some evidence suggesting that there's cost benefit calculation is going on in the minds of people in deciding whether to comply with state demands for information. They're concerned about, they're thinking about, you know, what do I get from this down the line, and also potentially what might I be exposed to in terms of taxation from a potentially predatory state.
[00:27:12] Alan Renwick: And as a result of that, you're saying the consequence of this system is to replicate and further entrench existing inequalities in society?
[00:27:20] Jeremy Bowles: That's right. Yeah.
[00:27:21] Alan Renwick: Yeah. So that is the core finding of the research.
Alas, we're coming towards the end of our time now.
But just quickly, before we wrap up, all of that was in the 1960s. How is this relevant today? Is this playing out in current policy debates about how things should be done?
[00:27:40] Jeremy Bowles: So to many extent, no, actually. Well, it's relevant, but it's not playing out.
And there are a few reasons for that.
So as I alluded to the initial interest for this project came from an interest in studying modern biometric schemes to identify and register citizens, which are often extremely sophisticated in very, very poor countries – much more so than they are in the UK or the US.
And at least in the last 10 years, there's been this real recognition that it's potentially a problem for governance. The fact that there are nearly like a billion people in the world who can't prove who they are to the state, to their government. They don't have any birth certificate or national ID or passport or anything like that.
But a lot of the solutions to this problem have been primarily a little bit technocratic. And so a lot of the international donor community that's involved in this, as well as private foundations and so on, have focused on this as being a little bit of almost like a technological problem of like, oh, if we make these clever technologies to enrol people, then we'll solve the problem rather than, you know, per se thinking about the local political economy of whether people actually want to comply with these things in the first place, or like what are the incentives facing citizens to actually comply with these things.
And there's often like an assumption today that these technologies sort of have a progressive incidence, so they're good for particularly poorer people. But I don't think that's obviously the case. At least, you know, this is historical evidence I'm showing, but this was a nominally kind of progressive state. But actually the impact of these schemes was quite regressive.
Just as today, it's sort of pointing to the fact that often citizens have quite mixed incentives on whether to comply.
And secondly, governments also face these different incentives in terms of rolling these schemes out in the first place and whether they necessarily face strong incentives to learn about the whole population.
And I, you know, I have a few other projects which try and get out that latter question a bit, to think about: Do governments actually face strong incentives to learn about all of their population or are these more truncated systems like I describe in Tanzania, are these in some sense reinforcing where actually they're quite happy just having informational capacity, as you might call it, covering relatively more economically elite people in the country versus expanding the sort of informational coverage to the whole population, because there might be some downsides of that?
And I think that these are kind of questions which are important to thinking about both why a lot of these modern technological schemes have failed, but also to thinking about what are the actual prospective welfare implications of government efforts to try and register everyone in fairly low capacity contexts.
[00:30:27] Alan Renwick: Well, thank you so much, Jeremy.
It's one of those, fascinating topics that sound really nerdy when you first look at them. You know, the role of information in state building. That sounds quite kind of dry. But actually it's incredibly important.
And one aspect of this general phenomenon that we often talk about and often hear about that, you know, you need to actually look at how things appear to people on the ground rather than just looking at how they appear to nerds like us or policymakers sitting in their offices.
So thank you fascinating, fascinating discussion.
We have been talking about Jeremy Bowles's article ‘Identifying the Rich: Registration, Taxation, and Access to the state in Tanzania’, which was published in this month's edition of the American Political Science Review.
Next week we will be exploring neocolonialism. Specifically we will be asking what can unequal trade patterns in cultural goods tell us about neocolonialism and what should – and can – be done about this.
Remember to make sure you don't miss out on that or other future episodes of UCL Uncovering Politics, all you need to do is subscribe. You can do so on Apple, Google Podcasts or whatever podcast provider you use. And while you're there, we'd love it if you could take a moment of time to rate or review us too.
I'm Alan Renwick. This episode was researched by Alice Hart and produced by Eleanor Kingwell-Banham. Our theme music is written and performed by John Mann.
This has been UCL Uncovering Politics. Thank you for listening.