The Lay of the Land: Information Capacity and the Modern State

• Author: Jan Teorell,Johannes Lindvall,Thomas Brambor,Agustín Goenaga
• DOI: 10.1177/0010414019843432
• Published date: 01 February 2020
• Date: 01 February 2020
• Subject Matter: Articles

https://doi.org/10.1177/0010414019843432

Comparative Political Studies

2020, Vol. 53(2) 175 –213

© The Author(s) 2019

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0010414019843432

journals.sagepub.com/home/cps

Article

The Lay of the Land:

Information Capacity

and the Modern State

Thomas Brambor1, Agustín Goenaga2,

Johannes Lindvall2, and Jan Teorell2

Abstract

This article presents new evidence on the efforts of states to collect and process

information about themselves, their territories, and their populations. We

compile data on five institutions and policies: the regular implementation of a

reliable census, the regular release of statistical yearbooks, the introduction

of civil and population registers, and the establishment of a government

agency tasked with processing statistical information. Using item response

theory methods, we generate an index of “information capacity” for 85 states

from 1789 to the present. We then ask how political regime changes have

influenced the development of information capacity over time. In contrast

with the literature on democracy and fiscal capacity, we find that suffrage

expansions are associated with higher information capacity, but increases in

the level of political competition are not. These findings demonstrate the

value of our new measure, because they suggest that different elements of

state capacity are shaped by different historical processes.

Keywords

information, political development, state capacity

When the term “statistics” was first used, in the 18th century, it referred to

scientific knowledge about the state. In this article, we present a statistical

analysis (in the current sense of the word) of statistical knowledge (in the

1Columbia University, New York, NY, USA

2Lund University, Sweden

Corresponding Author:

Johannes Lindvall, Lund University, Box 52, Lund, 22100, Sweden.

Email: johannes.lindvall@svet.lu.se

843432CPSXXX10.1177/0010414019843432Comparative Political StudiesBrambor et al.

research-article2019

176 Comparative Political Studies 53(2)

original sense of the word): We demonstrate when, where, and how states

began to gather and organize basic information about themselves, their terri-

tories, and the populations they govern.1

Our article contributes to the literature on state capacity, which is a matter

of increasing concern to economists, political scientists, and sociologists.

State capacity has been defined as “a government’s ability to make and enforce

rules, and to deliver services” (Fukuyama, 2013, p. 350), the “institutional

capability of the state to carry out various policies” (Besley & Persson, 2011,

p. 6), and the “degree of control that state agents exercise over persons, activi-

ties, and resources within their government’s territorial jurisdiction”

(McAdam, Tarrow, & Tilly, 2001, p. 78). Scholars typically treat state capacity

as “a function of state bureaucracy, the state’s relations with social actors, and

its spatial and societal reach” (Soifer & vom Hau, 2008, p. 220).

Among the resources states use when they “make and enforce” rules,

“carry out” their policies, and “control” persons, activities, and resources,

one of the most important is information. Our article, therefore, develops a

new measure of the information that states collect, store, retrieve, and process

about themselves. We rely on comparative, historical data on five informa-

tion-gathering and information-organizing activities: the introduction and

regular implementation of a national census, the introduction of civil regis-

ters the introduction of population registers, the establishment of a permanent

state agency tasked with processing statistical information about a country’s

territory and population, and the regular publication of statistical yearbooks.

We then analyze the empirical relationships among these institutions and pro-

grams with item response theory (IRT) methods, generating a combined

numerical measure of “information capacity.”

In the final part of the article, we examine the relationship between the

development of democratic institutions and the development of information

capacity from the 1790s onward. We find that suffrage expansions are associ-

ated with increases in information capacity, whereas changes in the level of

political competition are not. Interestingly, the development of fiscal capac-

ity, another important element of state capacity, has been shaped by different

historical processes. This demonstrates the value of our new measure and the

importance of distinguishing among the different resources that contribute to

higher state capacity.

Information and State Capacity

Scholars often infer the level of state capacity in a polity from data on policy

outputs and policy outcomes. For example, Soifer (2012), one of the world’s

leading scholars of the state, bases one of his measures of state capacity on

Brambor et al. 177

data on its expected outcomes, including low violent crime rates and low

private security expenditures (“coercive capability”), high vaccination rates

(“administrative capability”), and a high share of the population working in

the formal sector (“fiscal capability”). Hanson and Sigman (2013), to take

another prominent example, mix more direct evidence, such as survey-based

estimates of the efficiency of bureaucratic institutions, with outcome vari-

ables, such as data on tax evasion.2 In fact, there are few available measures

of state capacity that do not rely, at least in part, on data on policy outputs and

outcomes.

The drawback of this approach is that all arguments about the effects of

state capacity on policies and outcomes risk becoming tautological. It is pref-

erable, in our view, to concentrate on the “inputs” of state capacity—the

resources states use to enhance their capacity—and not the “outputs.”

At the most general level, the term “state capacity” refers to the ability of

a state to produce the outcomes political leaders attempt to achieve, whether

to “make and enforce” rules, to “carry out” policies, or to “control” popula-

tions and territories. To fix ideas, we can refer to these “attempts to achieve”

something as a vector of policies, p, and the intended “outcomes” as changes

in a vector of individual preferences or behaviors, y. Following Lindvall and

Teorell (2016), we can then think of state capacity as the strength of the

causal relationship between p and y: When the government of a high-capac-

ity state decides to adopt the policy p to achieve the outcome y, it is more

likely to be successful than the government of a low-capacity state would be,

if it adopted the same policy.

High-capacity states are better able to produce the outcomes their govern-

ments want to achieve because they deploy resources, denoted r, to increase

the likelihood that policies (p) achieve their intended outcomes (y). In other

words, the relationship between p and y is conditioned by r.

Because state capacity, in this framework, is a causal effect, it cannot be

measured directly; it can only be estimated. But the resources that states

deploy—the elements of r—can be measured. In our view, the most promis-

ing way forward for the literature on state capacity is, therefore, to measure

key resources, rather than measuring state capacity per se. “Key” resources

are fungible (transferable from one policy area to another), and their use is

independent of a particular state leader’s policy preferences. Institutions and

policies that enable states to collect, store, retrieve, and process reliable infor-

mation about “persons, activities, and resources within their government’s

territorial jurisdiction” (McAdam et al., 2001, p. 78) are crucial resources that

meet these two criteria.

Most of the things states do require information of some sort. Consider first

what Scott (1998) calls the “classic state functions of taxation, conscription