Is the Budget Deficit Sustainable when Fiscal Policy is nonlinear? The Case of Spain, 1961-2001

Abstract

[eng] Transportation costs and monopoly location in presence of regional disparities. . This article aims at analysing the impact of the level of transportation costs on the location choice of a monopolist. We consider two asymmetric regions. The heterogeneity of space lies in both regional incomes and population sizes: the first region is endowed with wide income spreads allocated among few consumers whereas the second one is highly populated however not as wealthy. Among the results, we show that a low transportation costs induces the firm to exploit size effects through locating in the most populated region. Moreover, a small transport cost decrease may induce a net welfare loss, thus allowing for regional development policies which do not rely on inter-regional transportation infrastructures. cost decrease may induce a net welfare loss, thus allowing for regional development policies which do not rely on inter-regional transportation infrastructures. [fre] Cet article d�veloppe une statique comparative de l'impact de diff�rents sc�narios d'investissement (projet d'infrastructure conduisant � une baisse mod�r�e ou � une forte baisse du co�t de transport inter-r�gional) sur le choix de localisation d'une entreprise en situation de monopole, au sein d'un espace int�gr� compos� de deux r�gions aux populations et revenus h�t�rog�nes. La premi�re r�gion, faiblement peupl�e, pr�sente de fortes disparit�s de revenus, tandis que la seconde, plus homog�ne en termes de revenu, repr�sente un march� potentiel plus �tendu. On montre que l'h�t�rog�n�it� des revenus constitue la force dominante du mod�le lorsque le sc�nario d'investissement privil�gi� par les politiques publiques conduit � des gains substantiels du point de vue du co�t de transport entre les deux r�gions. L'effet de richesse, lorsqu'il est associ� � une forte disparit� des revenus, n'incite pas l'entreprise � exploiter son pouvoir de march� au d�triment de la r�gion l

Full text

Page 1

Is the budget de…cit sustainable when
…scal policy is nonlinear? The case of
Spain, 1964-2001¤
Oscar Bajo-Rubioy
(Universidad de Castilla-La Mancha and centrA)
Carmen Díaz-Roldán
(Universidad de Castilla-La Mancha and centrA)
Vicente Esteve
(Universitat de València)
January 2004
Abstract
In this paper we re-examine the long-run sustainability of budget
de…cits, when …scal policy is conducted as a nonlinear process. Our
empirical methodology makes use of recent developments on threshold
cointegration that consider the possibility of a nonlinear relationship
between government expenditures and revenues. The analysis is ap-
plied to the case of Spain, a country that has recently accomplished an
important …scal consolidation. Overall, our results suggest the pres-
ence of signi…cant nonlinear e¤ects in Spanish …scal policy, so that …scal
authorities would cut de…cits only if they are ‘large’, and assuring their
long-run sustainability.
Keywords: Fiscal policy, Sustainability, Threshold cointegration,
Nonlinearity.
JEL classi…cation: E62, H62.
¤The authors wish two thank two anonymous referees for helpful comments to an earlier
version. Financial support from the Department of Science and Technology of the regional
government of Castilla-La Mancha, through the project PBI-03-003; as well as from the
Valencian Council of Education and Science, through the project GRUPOS03-151 (V.
Esteve), is also gratefully acknowledged.
yCorresponding author: Oscar Bajo-Rubio, Departamento de Economía y Empresa,
Universidad de Castilla-La Mancha, 13071 Ciudad Real (Spain). Tel.: +34-926-295300,
ext. 3580; fax: +34-926-295211. E-mail: oscar.bajo@uclm.es.
1

Page 2

1Introduction
The formation of the Economic and Monetary Union (EMU) by 12 member
countries of the European Union (EU) means that …scal policy becomes the
main instrument of stabilization policies available to their national authori-
ties. As a consequence, issues related to …scal policy have gained a growing
interest in recent years, in both academic and policy circles. In particular,
the long-run sustainability of budget de…cits has become a matter of great
concern. On the one hand, emphasizing the sustainability of de…cits would
mean in itself a limit to the ‡exibility of …scal policy as a stabilization tool;
on the other hand, EMU might lead to relax …scal discipline, so increasing
the risks of default and bailout. In turn, all this was re‡ected in the Treaty
of Maastricht, which de…ned budgetary rules that countries had to satisfy
in order to enter EMU, later enforced in the clauses of the Pact for Stability
and Growth (De Grauwe, 2003).
The traditional approach to the analysis of the sustainability of budget
de…cits has tested whether thegovernment’s intertemporalbudget constraint
(IBC) holds, that is, whether the current level of debt equals the present
discounted value of primary surpluses. Empirical tests on sustainability,
however, are still inconclusive due to di¤erences in the econometric method-
ology, the particular speci…cation of the transversality condition, and the
sample period used.
Several procedures to test for the IBC have been proposed in the lit-
erature, which focus on the univariate properties of the government de…cit
and debt (Hamilton and Flavin, 1986; Wilcox, 1989), and on the presence
of a long-run, linear, cointegration relationship between government rev-
enues and expenditures (Trehan and Walsh, 1988, 1991; Haug, 1991; Smith
and Zin, 1991). Further on, the eventual occurrence of structural breaks
in this cointegrating relationship has been examined by Hakkio and Rush
(1991a), who assumed the break point as exogenously given; and by Haug
(1995), Quintos (1995), Camarero, Esteve and Tamarit (1998), Makrydakis,
Tzavalis and Balfoussias (1999), and Martin (2000), where the break point
was endogenously derived.
However, and unlike the above quoted papers, which assumed a linear
relationship between government expenditures and revenues, several recent
studies have suggested the possibility that …scal policy may have non-linear
e¤ects, in the sense that both the size and the sign of the response of macro-
economic variables to …scal policy actions could be di¤erent depending on
the way and the initial conditions in which such policy actions are imple-
mented; see Giavazzi, JappelliandPagano (2000) for a recent comprehensive
empirical analysis.
The objective of this paper is to provide some additional evidence on
the sustainability of budget de…cits, when …scal policy is conducted as a
nonlinear process, which has been hardly treated in the literature. To this
2

Page 3

end, we make use of recent developments on threshold cointegration that
consider the possibility of a nonlinear relationship between government ex-
penditures and revenues. Our approach should be relevant for a number
of EU countries that have faced problems of …scal sustainability during the
1980s and 1990s. In particular, we analyze the case of Spain, a country
traditionally experiencing high budget de…cits, which has accomplished an
important …scal consolidation in last years that has allowed her to be able
to participate in EMU since the outset.
The rest of the paper is organized as follows. Some motivation for the
analysis in the paper, in terms of the recent literature on nonlinear …scal
policy, is provided in Section 2. Next, the empirical methodology is outlined
in Section 3. The results of the tests on threshold cointegration between
Spanish government expenditures and revenues are presented in Section 4.
Finally, the main conclusions are summarized in Section 5.
2 Nonlinearities in …scal policy
A wide attention has been recently given in the literature on stabilization
policy, to the possible non-Keynesian e¤ects of …scal policy and, in general,
its nonlinear e¤ects. The pioneering paper was Giavazzi and Pagano (1990),
who studied the cases of two strong …scal contractions performed during
the 1980s (namely, Denmark in 1983-86, and Ireland in 1987-89) that were
accompanied by a vigorous economic expansion, contrarily to what should
be expected according to a standard Keynesian view.
Such a result would be explained through the operation of two di¤erent
channels, i.e., the wealth channel and the expectations channel. On the one
hand, the …scal contraction can reduce interest rates, which would increase
the market value of non-human wealth, and so private consumption and
aggregate demand. On the other hand, a large enough …scal contraction
might lead the agents to expect lower taxes in the future, which would
increase households’ permanent disposable income, and hence their planned
and current consumption. Notice that this literature tends to stress the
role of the expectations channel, since the wealth channel is deemed to
be insu¢cient to generate the observed e¤ects on private consumption and
aggregate demand.
More generally, the possibility of non-Keynesiane¤ects followingchanges
in taxes or government spending would mean that …scal policy could behave
in a nonlinear way, so that both the magnitude and the size of its e¤ects
on macroeconomic variables would be di¤erent according to the particular
circumstances in which the …scal measures are enforced.
So, for instance, the non-Keynesian e¤ects would be more likely if …scal
policy changes are su¢ciently large and protracted (Giavazzi and Pagano,
1996). The composition of the …scal measures also matters, since cutting
3

Page 4

public sector wages and transfer programs, rather than raising taxes, would
tend to produce permanent reductions in the budget de…cit, which in turn
would favor the possibility of an expansionary e¤ect on aggregate demand
(Alesina and Perotti, 1995). And the latter, indeed, would be reinforced
when combined with two additional ingredients: some form of wage agree-
ment with the unions that ensures wage moderation, and a devaluation
immediately before the …scal tightening (Alesina and Ardagna, 1998).
Finally, the role of the initial conditions in which the …scal adjustment
is performed, also seems to be of great importance, with an expansionary
…scal contraction being more likely for high initial levels in tax rates or in
the ratio of debt to output (Blanchard, 1990). The latter point was later ex-
tended by Perotti (1999), who developed a model where …scal shocks would
have Keynesian e¤ects on private consumption (i.e., positive for government
expenditure shocks, negative for tax shocks) in “normal” times, and non-
Keynesian e¤ects in “bad” times, where “bad” times were de…ned in terms
of a high value of the accumulated government debt.
Until now we have examined under which circumstances a …scal policy
action, and in particular a …scal adjustment, could lead to e¤ects on macro-
economic variables that not always agree with those expected according to a
standard Keynesian view. But, what about the timing of the …scal actions?
As noticed by Alesina and Drazen (1991), although it could be agreed that
stabilization requires a change in …scal policy to eliminate budget de…cits,
stabilizations can be delayed if there is disagreement about how the bur-
den of the policy change is to be shared. Hence, stabilization would occur
only when certain groups give in, and allow the others to decide on how the
burden of the …scal adjustment is distributed.
More speci…cally, Bertola and Drazen (1993) argue that signi…cant cuts
in government spending take place only when the ratio of government spend-
ing to output hits a trigger point, which implies that abrupt changes in …scal
policy should not be observed frequently. In other terms, these authors in-
troduce a nonlinearity in the reaction function of …scal policymakers. And
this in turn would imply a nonlinear relationship between the consumption-
to-output ratio and the government spending-to-output ratio, provided that
the current policy generates su¢ciently strong expectations of future policy
changes in the opposite direction.
The above considerations should be of special relevance for the case of
EMU. As is well known, those countries wishing to participate in EMU
had to satisfy, according to the Treaty of Maastricht, several “convergence
criteria”, two of them related to …scal policy; namely, the budget de…cit and
government debt should not exceed 3% and 60%, respectively, of their GDP.
Leaving aside the suitability of these …scal criteria, whose economic rationale
was questioned by several authors [see, e.g., Buiter, Corsetti and Roubini
(1993) or De Grauwe (1996)], there was at that time a wide discussion
about the potential di¢culty of ful…lling them for most of the countries
4

Page 5

involved. So, for instance, Wyplosz (1997) observed that, given the tight
monetary policies performed during the mid-1990s to meet the criterion on
in‡ation, growth slowed down, which reduced tax revenues and made even
more di¢cult to achieve the …scal targets. However, most countries were
able in the end to satisfy the …scal and other convergence criteria, and so to
participate in EMU from the start. What could it happen?
The objective of this paper, then, will be trying to quantify the pre-
cise moment in which the authorities would correct a de…cit, so that a
required …scal adjustment is …nally enforced; i.e., the “trigger point”, ac-
cording to Bertola and Drazen’s (1993) contribution. Notice, on the other
hand, that the empirical literature on …scal sustainability has hardly incor-
porated the new developments on nonlinearities in …scal policy. We can
just quote Cipollini (2001), who introduced a regime shift in the adjustment
towards a linear long-run (cointegrating) relationship between total govern-
ment revenues and expenditures for the UK, using a smooth transition error
correction model to test for nonlinearities or asymmetries in the adjustment
process.
More recently, Bajo-Rubio, Díaz-Roldán and Esteve (2004) have found
strong evidence of nonlinearities in the evolution of the Spanish budget
de…cit in terms of a threshold autoregressive model, so that the de…cit dy-
namics would be di¤erent depending on whether the change in the de…cit
was below or above an endogenously estimated threshold; in other words,
mean-reverting dynamic behavior in the budget de…cit would be expected
once such threshold was reached. In particular, signi…cant …scal stabiliza-
tions were found to occur when, in a certain year, the ratio of the budget
de…cit to GDP showed an increase of more than 1.9% between the previous
year and the sixth year before.
Unlike our previous paper, which made use of a threshold autoregres-
sive model for the budget de…cit, we analyze here the possible presence of
threshold cointegration between government expenditures and revenues. In
other words, our objective will be analyzing the evolution of government
expenditures and revenues, rather than the budget de…cit as a whole.
3 Methodology
The concept of threshold cointegration was introduced by Balke and Fomby
(1997) as a feasible way to combine nonlinearity and cointegration. As is
well known, systems in which variables are cointegrated can be character-
ized by an error correction model (ECM), which describes how the variables
respond to deviations from the equilibrium. In this way, the ECM can be
characterized as the adjustment process through which the long-run equilib-
rium is maintained. The traditional approach, however, assumes that such
a tendency to move towards the long-run equilibrium is present every time
5