Explaining fiscal decentralization and the role of ethnic Diversity
| Author | Harry Pickard |
| Published date | 01 November 2020 |
| DOI | http://doi.org/10.1111/sjpe.12246 |
| Date | 01 November 2020 |
Scott J Polit Econ . 2020;67:469–485.
|
469
wileyonlinelibrary.com/journal/sjpe
1 | INTRODUCTION
Decentraliz ation has been linked to numero us economic outcomes. For in stance, a higher degree of f iscal decen-
tralization is a ssociated with less corrupt ion (Fisman & Gatti, 2002), a sm aller informal sector (Teobal delli, 2010)
and a more effi cient provision of public goods (E scaleras & Register, 2010; Faguet, 20 04). More decentralizati on
has been shown to re duce the effectivenes s of aid (Lessmann & Markwar dt, 2016) and to lower the rate of infla-
tion (Baskara n, 2012; Neyapti, 2004). Acros s countries, we observe ver y different levels of decentr alization and
there is a small por tion of literature that s eeks to address how this va riation has come about. M oreover, given the
Accepted: 27 March 2020
DOI: 10 .1111/sjpe.1 2246
ORIGINAL ARTICLE
Explaining fiscal decentralization and the role of
ethnic Diversity
Harry Pickard
This is an open ac cess article und er the terms of the Crea tive Commons Attr ibution License, w hich permits use , distribution an d
reproduct ion in any medium, pro vided the origina l work is properly cit ed.
© 2020 The Autho rs. Scottish Jou rnal of Political Eco nomy published by Joh n Wiley & Sons Ltd on beha lf of Scottish Econom ic
Society
Departm ent of Economics, Uni versity of
Sheffield, Sheffield, UK
Correspondence
Harry Pickard, Department of Economics,
Universit y of Sheffield, 9 Ma ppin Street,
Sheffiel d, S1 4DT, UK.
Email: h.g.pickard@sheffield.ac.uk
Abstract
This paper consider s the causes of fisc al decentralizati on
with a specific focu s on the role of ethnic diversity. To do so,
I employ an instrum ent for ethnic diversity based on t he ori-
gin of anatomically mod ern human life. Using two measures
of decentralization t hat capture decision m aking autonomy
and accounting for the d epth of divisions bet ween ethno-
linguistic groups us ing the structure of language t rees, I find
that ethnic diversit y has a positive effe ct on the degree of
decentralization . It is the amount of frac tionalization to-
wards the leaves of the tre es, where groups ar e more nu-
merous and less distinct, that drive decentralization.
KEYWORDS
ethno-linguistic diversity, fractionalization, instrumental variable
JEL CLASSIFI CATION
H110; H710; H 770; Z10
[Correcti on added on 03 Sep tember 2020, af ter the first o nline public ation: The art icle categor y has been update d in this version]
470
|
PICKARD
variety of outco mes decentra lization is associ ated with, it is cri tical to under stand how count ries have become
relatively more decentralized than others.
The overarching o bjective of this paper is to e xplore the determinant s of fiscal decentrali zation with a specific
focus on the role of et hno-linguist ic fraction alization. Th e relationship has b een covered in the li terature and
has been shown to be p ositively linked to decentraliza tion, and has also b een shown to affect a wide variet y of
economic and po litical outcomes. These t ypes of fractionaliz ation measure were first intr oduced by Easterly and
Levine (1997), who show the ne gative consequences of ethni c fractionalization on dev elopment, which has since
been confirme d by Alesina, Dev leeschauwer, Eas terly, Kurlat, and Wac ziarg (2003), De smet, Ort uño-Ortí n, and
Wacziarg (2012), As hraf and Galor (2013b), Duch (2 014) and Papyrakis and Mo (2014). Addit ionally, others demon-
strate how inter nal fractionalizati on can increase the risk of t he onset of violent civil confli ct (Desmet et al., 2012;
Fearon & Laitin , 2003; Montalvo & Reynal -Querol, 2005).
The contribut ion of this paper lies in the tre atment of fractionalizat ion as an endogenous variab le. The previ-
ous work assumes f ractionali zation as exogenou s, however, this assumpti on is being challe nged in recent work
(Casey & Owen, 2014; Dari ty, Mason, & Stewart, 20 06). As most migrat ion occurs into those cou ntries with higher
levels of economic a nd institutional develo pment, both positively as sociated with decentral ization, it implies that
OLS coeffi cients will be dow nwardly biased (Fre eman, 2006). Reve rse causalit y and omitte d variables are a lso
likely a source of end ogeneity in terms of decentr alization. In light of these co ncerns, I treat diversit y as endoge-
nous to decentra lization. I follow the previ ous work by Ahlerup and Olss on (2012) on the historical de terminants
of fractionalization, which identifies a suitable instrument for examining the role ethno-linguistic heterogeneity.1
Specifica lly, I use the origin of anatomica lly modern human life in e ach country as an inst rument for ethno-ling uis-
tic diversit y. This follows from the premis e that a longer settlem ent duration breeds mo re ethnic heterogeneit y. In
terms of data on fr actionalization, I us e the Desmet et al. (2012) measur es that account for the depth of d ivisions
between ethn o-linguistic groups. T his allows for a distinctio n in the depth of the divisions b etween groups in the
analysis and is als o new to the decentralizat ion literature. To capture the d egree of decentralizat ion I employ data
produced by Hoog he et al. (2016), the Regional Aut hority Index (R AI). These data are n ot new (Hooghe et al., 2016;
Niedzwieck i, Osterkatz, Hoogh e, & Marks, 2018), although t his paper seeks to expla in the variation in two par tic-
ular component s of the RAI that captu re a truer degree of fisca l decentralization than t he traditional measu res.
This paper fin ds that there is a posi tive causal rel ationship bet ween the level of eth no-linguisti c fractional -
ization and the d egree of decentr alization. Mor e specificall y, it is the amount of frac tionalizati on where there
are more groups , which are less distinctl y defined, that drive the dece ntralizing of fiscal autho rity. These are the
groups that are fo rmed from the leg acy of coloniali sm and uninterru pted human sett lement. Whe reas, the pe-
rennial groups , where cleavages are at their d eepest, bear no impac t on the degree of decentral ization. Crucially,
when treating f ractionalization as a n endogenous variable in t he appropriate manner, the rel ationship persists .
The rest of this pa per is organized as follows. The p roceeding section prese nts and discusses the theo retical
determinant s of decentralization that h ave been used in the previous lit erature. Section 3 des cribes the data and
presents th e empirical strateg y. Section 4 present s the results of the emp irical analysis. Se ction 5 presents a nu m-
ber of robustne ss checks on the result s and Section 6 conclude s.
2 | THEORETICAL CONSIDERATIONS
2.1 | Ethno-linguistic heterogeneity
The primar y focus of this paper is the role of eth no-linguistic diversit y, which has a stron g theoretical grounding
with respect to decentralization—first introduced into the decentralization theorem by Oates (1972). There are at
1There are othe r seminal works o n the origins of et hno-linguis tic diversit y (see, e.g., Mich alopoulos (20 12) and Ashraf an d Galor (2013a)).
To continue reading
Request your trial