Distribution or Diversion? Distribution of Local Public Goods in the Presence of Clientelist Brokers

• Date: 01 June 2021
• DOI: 10.1177/1065912920903800
• Published date: 01 June 2021
• Subject Matter: Articles

https://doi.org/10.1177/1065912920903800

Political Research Quarterly

2021, Vol. 74(2) 274 –287

© 2020 University of Utah

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1065912920903800

journals.sagepub.com/home/prq

Article

Existing literature finds that leaders in developing coun-

tries often—though not always—allocate more local pub-

lic goods (LPGs) to the communities that vote for them

(Albertus 2013; Rosas, Johnston, and Hawkins 2014;

Vaishnav and Sircar 2012). In most cases, these patterns

are interpreted as distribution from the center to voters in

exchange for their support; a model in which voters sup-

port politicians in anticipation of better LPGs is particu-

larly prominent in the African politics literature (Carlson

2015; Ichino and Nathan 2013; Posner 2005).

But this conclusion sits uneasily with other literature

that suggests that blocs of voters in poor countries do not

directly negotiate a quid-pro-quo with the central state

but are delivered to the regime by networks of local bro-

kers who either compel political loyalty or buy it with

targeted private goods (Baland and Robinson 2008;

Boone 2011; de Kadt and Larreguy 2018; Frye, Reuter,

and Szakonyi 2014; Koter 2013). Clientelist brokers

require access to discretionary funds for vote-buying and

their own compensation (Camp 2017; Camp and

Szwarcberg 2015; de la O 2015; Jablonski, Sacks, and

Larizza 2014; Kasara 2007; Novaes 2015; Singer 2009;

Trantidis and Tsagkroni 2017; Zarazaga 2015), but since

governments can’t overtly allocate money to clientelism,

the money must be diverted from other components of the

budget. This raises the question: When we see that loyal

constituencies are granted allocations for LPGs, are

voters actually receiving these goods? Or are these funds

intended for use by the local brokers who keep voters

loyal?

The question is relevant to our models of how voters

and leaders behave. Most distributive politics literature

assumes two actors: voters, who vote in response to dis-

tribution, and elected leaders, who distribute goods to

earn support. If this is the correct model, then the correla-

tion between local electoral outcomes and allocations to

LPGs will be sufficient to determine whether there is an

exchange of goods for support. In clientelist countries,

however, there will likely be another actor, an intermedi-

ary (or coordinated group of intermediaries) who can

both divert resources and influence vote choice, and the

intermediary’s actions will moderate the apparent rela-

tionship between vote choice, official allocations, and de

facto LPGs provision. Communities who appear to sup-

port the incumbent in exchange for greater allocations of

LPGs may actually be voting in exchange for handouts

bought with diverted funds. Conversely, if we find a neg-

ative correlation between observed LPGs and support for

903800PRQXXX10.1177/1065912920903800Political Research QuarterlyCarlson

research-article2020

1Penn State, University Park, PA, USA

Corresponding Author:

Elizabeth Carlson, Penn State, 228 Pond Lab, University Park, PA

16802, USA.

Email: ecc13@psu.edu

Distribution or Diversion? Distribution

of Local Public Goods in the Presence

of Clientelist Brokers

Elizabeth Carlson1

Abstract

In this article, I argue that, in clientelist regimes, money officially allocated to local public goods will sometimes

be diverted to private handouts, and that the extent of diversion will be systematically correlated with electoral

outcomes. I show that allocated funds are more likely to translate into local public goods in opposition areas and the

incumbent’s ethnic core. Diversion of funds is greatest in areas outside the ethnic core that are nevertheless won

by the regime. This is consistent with a model in which brokers tasked with delivering non-coethnic voters to the

regime divert funds for this purpose. As a result of extensive diversion in loyal non-coethnic districts, these voters are

significantly less likely to see completed local public goods projects than either core voters or those who oppose the

regime. This unexpected result suggests that the behavior of local brokers may be responsible for a number of puzzling

findings in the distributive politics literature.

Keywords

clientelism, distributive politics, African politics