9738

2022
May 2022

Does Government Spending

Efficiency Improve Fiscal
Sustainability?
António Afonso, José Alves
Impressum:

CESifo Working Papers

ISSN 2364-1428 (electronic version)
Publisher and distributor: Munich Society for the Promotion of Economic Research - CESifo
GmbH
The international platform of Ludwigs-Maximilians University’s Center for Economic Studies
and the ifo Institute
Poschingerstr. 5, 81679 Munich, Germany
Telephone +49 (0)89 2180-2740, Telefax +49 (0)89 2180-17845, email office@cesifo.de
Editor: Clemens Fuest
https://www.cesifo.org/en/wp
An electronic version of the paper may be downloaded
· from the SSRN website: www.SSRN.com
· from the RePEc website: www.RePEc.org
· from the CESifo website: https://www.cesifo.org/en/wp
 CESifo Working Paper No. 9738

Does Government Spending Efficiency

Improve Fiscal Sustainability?

Abstract

We evaluate the impact of government spending efficiency on fiscal sustainability for a panel of
35 OECD countries during the period of 2007-2020. To answer our research question we first
compute the magnitude of the responses of government revenues to changes in government
spending. Next, we make use of so-called government spending efficiency scores, which
efficiently indicate how governments can maintain their level of performance whilst using fewer
inputs. Our results show that for the input efficiency scores obtained, countries’ fiscal balance and
fiscal sustainability is directly improved by the use of less public resources, whilst maintaining
the same level of output. In the cases of the output efficiency scores, the commitment of increased
government outputs can lead to higher economic growth and the generation of additional
government revenues, which also improves fiscal sustainability. Specifically, rationalising public
expenditures without jeopardising the actual level of public goods and provision of services is a
stronger determinant of fiscal sustainability, as well as for the improvement of the primary budget
balance.
JEL-Codes: C230, E210, E620, H500, H620.
Keywords: fiscal sustainability, spending efficiency, panel data.

António Afonso José Alves

ISEG – Lisbon School of Economics and ISEG – Lisbon School of Economics and
Management, Universidade de Lisboa Management, Universidade de Lisboa
Lisbon / Portugal Lisbon / Portugal
aafonso@iseg.ulisboa.pt jalves@iseg.ulisboa.pt

April 2022
This work was supported by the FCT (Fundação para a Ciência e a Tecnologia) [grant number
UIDB/05069/2020]. The opinions expressed herein are those of the authors and do not necessarily
reflect those of the authors’ employers. Any remaining errors are the authors’ sole responsibility.
1. Introduction
An ongoing macro and fiscal discussion pertains to the issue of whether public finances
follow a sustainable intertemporal path. This question becomes even more relevant when
economies are hit by diverse unforeseen shocks (e.g., financial, health, or energy crises, or war-
related fallout developments), which usually prompt governments to take the necessary
responses. Such policy measures generally imply incurring more government spending and lead
to higher fiscal imbalances.
Specifically, the financial situation of public accounts at the start of 2021 in most
advanced economies was much more difficult, both in terms of debt-to-GDP ratio and primary
budget balances. Such difficulties may even have become structural, contrary to some views
that the effect of COVID-19 on public accounts is purely cyclical, which illustrates how serious
the effect of the unanticipated shocks can be on fiscal sustainability
Typically, fiscal sustainability is considered to be in place when government revenues
move closely in pace with government spending and when the government is able to deliver a
primary balance that is sufficient to stabilise the debt-to-GDP ratio. Furthermore, fiscal
sustainability is linked (influenced) to (by) certain key factors, such as, for instance, sovereign
debt behaviour, business cycle developments, the cost of long-term sovereign borrowing, and
fiscal rules. In addition, the substantial outstanding stock of government debt, notably in many
OECD economies, has especially fomented debate about the possibility of joint debt
management at the Euro Area level (see, for instance, Amato and Saraceno, 2022).
Nevertheless, it can also be hypothesised that the efficient use of public resources by
governments can be (should be) growth enhancing and, that this therefore effectively
contributes to the increased sustainability of fiscal policies. Accordingly, in this paper we
combine two topics and strands of literature, namely: fiscal sustainability and government
spending efficiency. We proceed to postulate the research premise that the efficiency of
government spending could have a positive effect on fiscal sustainability. In fact, a more
efficient use of government spending, targeted to growth-enhancing activities, and with little
wasted resources, would likely guarantee higher macroeconomic rates of return and avoid the
undue crowding out of private investment (see, Afonso and St. Aubyn, 2019) and foster growth
and deter fiscal imbalances.
Accordingly, in a context of unprecedented fiscal challenges for many economies, and
in order to assess the research question, we first compute the magnitude of the responses of
government revenues to changes in government spending for a panel of 35 OECD countries
during the period of 2007-2020 in order to test the hypothesis that both sides of the budget

2
balance should co-move. Second, we make use of so-called government spending efficiency
scores, which demonstrate notably how governments can increase their performance whilst
maintaining the same level of inputs, and also how governments can reduce the level of inputs
while maintaining the same level of performance. In this case, we use three different models to
obtain the efficiency scores, both output-oriented and input-oriented. Third, we empirically
evaluate the responsiveness of fiscal sustainability to changes in government spending
efficiency.
With regards the answer to the title question, our results show notably that a more
efficient government does indeed contribute to increased fiscal sustainability. In the case of the
input efficiency scores, the underlying rational implies using less public resources to maintain
the same level of output, which in turn directly improves both the fiscal balance and fiscal
sustainability. In the case of the output efficiency scores, the explanation for the results obtained
can be explained by the provision of more and better government outputs, which contribute to
higher levels of economic growth and added government revenues, which in turn improve fiscal
sustainability. More specifically, rationalising public expenditures without jeopardising the
actual level of public goods and the provision of services is found to be a better determinant of
fiscal sustainability than improving the primary budget balance.
The paper is organised as follows. Section 2 reviews the two-subject related literature.
Section 3 discusses the methodology. Section 4 presents the data and the analysis of the results,
and Section 5 is the conclusion.

2. Literature
2.1 Fiscal sustainability
In the context of fiscal sustainability, the studies of Hamilton and Flavin (1986) and
Trehan and Walsh (1991) applied to the United States are pioneers. Hakkio and Rush (1991)
sustain that when government revenues and expenditures series are non-stationary, then the
existence of cointegration between both variables is a necessary condition for the government
to comply with current value budget constraints.
For instance, Vanhorebeek and Rompuy (1995), Papadopoulos and Sidiropoulos (1999)
and Afonso (2005) research the fiscal solvency of several European Union countries.
Furthermore, the related literature has notably assessed the long-term relationship between
public revenues and expenditures, notably for advanced economies, concluding that the
sustainability of fiscal policy does indeed exist (Afonso and Rault, 2015; Afonso and Jalles,
2017; Brady and Magazzino, 2018, Magazzino et al, 2019). On the other hand, fiscal reaction

3
functions have also been used in research for estimating the response of primary balances to the
development of government debt, in line with Bohn (2008).
Among the possible factors that contribute to sustainable fiscal developments, the cost
of long-term sovereign borrowing vis-à-vis the economic growth rate also plays a role. For
instance, Blanchard et al. (2020) debate the relevance of the interest rate-growth differential in
the context of fiscal reaction functions. In the same vein, Afonso et al. (2022a) report that the
interest rate (r)-growth rate (g) differential matters. Indeed, for 28 EU countries over the period
of 1995 Q1 to 2021Q2, they discover a higher magnitude for the decrease in the debt-to-GDP
ratio following improvements in the primary balance when a positive interest rate-growth rate
differential exists. In addition, the business cycle can decisively influence fiscal policy, which
is important for the need to stabilise more resilient fiscal policies to business cycles. As
demonstrated in Aldama and Creel (2021), the response of fiscal policy to business cycle tends
to be asymmetric, i.e., it is found to be pro-cyclical in downturns, while fiscal policy is not
sensitive to economic upturns.
In addition, the potential lack of fiscal sustainability also raises the issue of the
interactions between fiscal policy and monetary policy. Indeed, if the primary budget balance
is set independently of observed debt levels, this could “force” monetary policy to adjust (a
passive behaviour) in order to guarantee the fulfilment of intertemporal government budget
constraints, with the price level also being influenced by fiscal developments (see Leeper, 1991;
Sims, 1994; and Woodford, 1995 for the discussion of the Fiscal Theory of the Price Level
hypothesis).

2.2. Public sector efficiency

The relevance of public sector efficiency has become a topic of growing interest in the
literature (see, for example, the works by Gupta and Verhoeven, 2001; Afonso et al., 2005,
2010). Several studies assess the degree of efficiency of the public sector by looking at different
country samples and time spans, using DEA and semi-parametric approaches, however most
tend to focus on OECD and European countries (Adam at al., 2011; Dutu and Sicari, 2020;
Afonso and Kazemi, 2017; Antonelli and de Bonis, 2019).
In the majority of these studies, the results point to the existence of possible public
sector efficiency gains. For instance, Afonso et al. (2005) report that the average input
efficiency in 2000 for 23 OECD countries is 0.79, which means that these countries should be
able to attain the same level output while using only 79% of the inputs that they currently use.
Similar results were reported for other country sets, both globally and for sectoral (social)
4
spending, with more recent analysis, respectively by Afonso and Kazemi (2017) and Antonelli
and de Bonis (2019).
Nevertheless, some studies have also addressed overall government spending efficiency
for African countries (Gupta and Verhoeven, 2001), Emerging Markets (Afonso et al., 2010),
Latin American countries (Afonso et al., 2013), Indian states (Mohanty et al, 2022), and Sub-
Sahara Africa (Olanubi and Olanubi, 2022), with all using similar non-parametric frameworks.
All the above-mentioned studies identified substantial public spending efficiency
differences between countries, as well as scope for savings in expenditure, which suggests that
government spending efficiency could be improved. This typically implies that more public
services could be provided, while using the same amount of public resources, or conversely,
that the same level of public resources could be provided with fewer public resources. Hence,
fiscal improvements in this respect can be positive for better assessments of financial markets.
In addition, to explain these cross-country efficiency differences, other studies have
examined factors such as: population, education, income level, quality of institutions (property
right security and level of corruption), quality of the country’s governance, government size,
governments’ political orientation, voter participation rate, and civil service competence
(Afonso et al., 2005; Hauner and Kyobe, 2010; Antonelli and de Bonis, 2019). More recently,
Afonso et al. (2021) evaluated the role of tax structures and tax reforms in explaining cross-
country efficiency differences. Furthermore, fiscal rules and government spending efficiency
are found to be important for explaining fiscal sustainability, albeit, these institutional variables
are considered to be substitutes, as fiscal rules tend to be the main explanation for higher fiscal
sustainability when government efficiency is on the increase (Bergman et al., 2016).

3. Methodology
With regards fiscal sustainability, we follow Afonso (2005) and Afonso and Jalles
(2017), mainly to assess whether a linear combination of government revenues and government
expenditures is stationary. If that is the case, then government revenues and expenditures
become cointegrated, which implies that the variables are attracted to a stable long-run
(equilibrium) relation with only short-run (temporary) deviations from the equilibrium.
Therefore, we estimate the following regression for each country1:

1
Note that the issue of stationarity of variables has been assessed. In table A2 in the Appendix we checked country-
by-country the stationarity properties of both government revenues and expenditures, for both levels and first
differences. The majority of countries have non-stationary variables (one or the two). Additionally, we tested the
cointegration between both revenues and expenditures series for each country, and discovered a long-term relation
between these two fiscal variables for all the countries in our sample.

5
 𝑅𝑡 = 𝛼 + 𝛽𝐺𝑡 + 𝑢𝑡 (1)

where 𝑅𝑡 denotes the government revenue-to-GDP ratio, and 𝐺𝑡 denotes the government
expenditure-to-GDP ratio. 𝑢𝑡 is a standard i.i.d. disturbance term that satisfies all the usual
assumptions. The closer to unity the estimated  coefficient is, the more sustainable public
finances will be, and a unitary increase in government spending will be matched by a  increase
in government revenues. In addition, we estimate Equation (1) by using an expanding widow
approach. That is to say, in order to obtain a time-series of 𝛽 for each country, we estimate
Equation (1) between 1980 and 2007; 1980 and 2008; 1980 and 2009;…; and lastly, for the
period of 1980-2020 for each country considered in our study, respectively.
Second, the other main variable of interest arises from the measure of public sector
efficiency. In this case, we use the so-called public sector efficiency scores as computed by
Afonso et al. (2022b). These public sector efficiency scores are computed by using data
envelopment analysis (DEA).2 This is deemed to be a suitable approach for several reasons:
first, it does not impose an underlying production function, and second, it accommodates
deviations from the efficient frontier and also examines the efficiency of a country in relation
to its peers.
(2) illustrates the case of the use of an input-oriented approach to measure the
proportional reduction in inputs, while holding the output constant. One also assumes variable-
returns to scale (VRS) to account for the fact that countries may not operate at the optimal scale.
Conversely, from an output-oriented perspective, one can assess how much output could
increase if the same level of inputs was maintained. The efficiency scores are computed by
applying the following linear programming problem3:

min 𝜃
𝜃,𝜆

𝑠. 𝑡. − 𝑦𝑖 + 𝑌𝜆 ≥ 0
𝜃𝑥𝑖 − 𝑋𝜆 ≥ 0 (2)

𝐼1’𝜆 = 1
𝜆≥ 0

2
DEA is a non-parametric frontier methodology, drawing from Farrell’s (1957) seminal work, which was further
developed by Charnes et al. (1978). Coelli et al. (2002) offer an introduction to DEA.
3
This is the equivalent envelopment form (see Charnes et al., 1978), which uses the duality property of the
multiplier form of the original model.

6
where 𝑦𝑖 is a column vector of outputs, 𝑥𝑖 is a column vector of inputs, 𝜃 is the efficiency
scores, 𝜆 is a vector of constants, 𝐼1’ is a vector of ones, 𝑋 is the input matrix, and 𝑌 is the
output matrix.
In Equation (2), 𝜃 is a scalar (which satisfies 0 ≤ 𝜃 ≤ 1) that measures the distance
between a country and the efficiency frontier. The efficiency frontier is defined as being a linear
combination of the best sampled countries (but not necessarily the best possible one). If 𝜃 < 1,
then the country is within the frontier and it is inefficient, whereas if 𝜃 = 1, this implies that
the country is on the frontier and that it is efficient.
We used three different DEA models, namely: the baseline model (Model 0), which
includes just one input (government spending as percentage of GDP) and one output, and is in
effect a composite public sector performance (PSP) indicator; Model 1 includes two inputs,
governments’ normalised spending on opportunity and on “Musgravian” indicators and one
output, with total PSP scores; finally, Model 2 uses one input, governments’ normalized total
expenditure, and two outputs, the opportunity PSP and the “Musgravian” PSP scores4. Detailed
results and definitions are reported in Afonso et al. (2021).5
We are thus interested in uncovering a positive contribution between government
spending efficiency and fiscal sustainability. Our testing specification hypothesis is as follows:

𝛽𝑖𝑡 = 𝛼1 + 𝛼2 𝛽𝑖𝑡−1 + 𝛼3 𝜃𝑖𝑡 + 𝛼4 𝜃𝑖𝑡 𝐷𝑖𝑡 + 𝜋𝑋𝑖𝑡 + 𝑢𝑖𝑡 (3)

where β is the time-varying sustainability coefficient, 𝜃 is the efficiency score obtained from
(2), D is a dummy variable to test for the relevance of the magnitude of the efficiency score (for
instance, efficiency level above 0.25 or if the country belongs to Eurozone), and X is a set of
other relevant sustainability explanatory factors. Such factors can include notably: the primary
balance (pbalance), the change in the level of sovereign indebtedness (Δdebt), the output gap

4
We present the correlation matrices for both input-oriented and output-oriented models in Tables A6 and A7 in
the Appendix, respectively.
5
Afonso et al. (2005) used a set of metrics to construct a composite public sector performance (PSP) indicator.
PSP is the simple average between so-called opportunity and Musgravian indicators. The opportunity indicators
evaluate the performance of the government in terms of administration, education, health and infrastructure
sectors, with equal weighting. The Musgravian indicators include three sub-indicators: distribution, stability, and
economic performance, all of which also have equal weighting for the indicators. Accordingly, the opportunity
and Musgravian indicators result from the average of the measures included in each sub-indicator. To ensure a
convenient benchmark, each sub-indicator measure is first normalised by dividing the value of a specific country
by the average of that measure for all the countries in the sample.

7
(outputgap), and, finally, the differential between the interest rate (r) and the output growth rate
(g).
We estimate (3) in a panel setup due to: the fact that we can use the information
contained in the cross-section dimension and increase the performance and accuracy of the
tests; the existence of cross-country dependence which can mirror common changes in the
behaviour of fiscal authorities (e.g., capital markets views, sovereign rating grouping, increased
business cycle synchronisation, peer pressure, and Euro Area grouping); and common policy
shocks, which can affect fiscal positions in several countries where policies and trade are more
interconnected.
In addition, we estimate Equation (3) using OLS-Fixed Effects 2SLS estimator in order
to compensate for endogeneity issues, as well as Weighted Least Squares with Fixed-Effects
(WLS-FE). Indeed, since our dependent variable is based on estimates, the error 𝑢𝑡 in (1) and
σ2
(3) is distributed as 𝑢𝑡 ∼ N (0, 𝑠 ), where 𝑠𝑖 are the estimated standard deviations of the
𝑖

time-varying sustainability coefficients for country i, and σ2 is an unknown parameter that

is estimated in the second-stage regression.

4. Empirical assessment
4.1. Data
Our analysis covers a panel of 35 OECD economies during the period of 2007-2020.
The country sample is as follows: Australia, Austria, Belgium, Canada, Chile, Colombia, Czech
Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy,
Japan, Latvia, Lithuania, Luxembourg, the Netherlands, New Zealand, Norway, Poland,
Portugal, Slovakia, Slovenia, South Korea, Spain, Sweden, Switzerland, Turkey, the United
Kingdom, and the United States.
This time span is that for which data to calculate the efficiency scores is available (see
Afonso et al., 2022b). Nevertheless, and in order to compute the sustainability magnitude, we
used data starting in 1980 for the respective expanding window. The detailed information
regarding the data used in the analysis is reported in the Appendix.
A first cursory look at the two main variables shows the magnitude of the sustainability
coefficient, while the efficiency scores shows the expected pattern of co-movement. Figure 1
shows stylised evidence for some countries, using the DEA output-oriented Model 2 (one input,
governments’ normalized total expenditure, and two outputs, the opportunity PSP, and the
“Musgravian” PSP scores).

8
 For instance, in the case of the United Kingdom, an average output efficiency score of
around 0.78 for the period of 2007-2022 can be observed, which highlights the possibility of
theoretically obtaining 22% more in terms of outputs with the same level of input (government
spending). On the other hand, during the same time period, the average value for the 
coefficient in the UK is around 0.34, which is not close to unity, which implies that government
revenue developments were lagging behind the stronger growth dynamics of government
spending (a fiscal sustainability issue).

Figure 1 – sustainability magnitude (RHS) and efficiency scores (LHS)

a b
Germany Portugal
0.94 0.25 0.9 0.6
0.92 0.2 0.5
0.85
0.9 0.15 0.4
0.88 0.1 0.8 0.3
0.86 0.05 0.2
0.75
0.84 0 0.1
0.7 0
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020

2007 2009 2011 2013 2015 2017 2019

teout2n beta teout2n beta

c d
Turkey United Kingdom
0.85 0.85 0.92 0.5
0.8 0.8 0.9 0.4
0.75 0.75
0.88
0.7 0.7 0.3
0.65 0.65 0.86
0.2
0.6 0.6 0.84
0.55 0.55 0.82 0.1
0.5 0.5 0.8 0
2020
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019

2007 2009 2011 2013 2015 2017 2019

teout2n beta teout2n beta

Note: teout2n (: Model 2, output-oriented efficiency score (one input, governments’ normalised total expenditure
and two outputs).

4.2. Results
4.2.1. Input-Oriented Efficiency

We first estimate the impact of public spending efficiency on fiscal sustainability,

employing the set of input-oriented efficiency scores. We present the results of Models 1 and 2

9
in Tables 1 and 2, respectively.6 Both tables present three different estimation results: including
one for the whole dataset; another one for the country-year pairwise whose correlation between
the efficiency scores and the sustainability magnitudes is above 0.25 (to focus on the cases
where a closer nexus can expected). The results obtained lead us to conclude that higher public
spending efficiency positively contributes to improve fiscal sustainability. Indeed, we have
obtained an expected and statistical positive coefficient for θ, not only when we estimate
Equation (3) without control variables, but also when we include fiscal and macroeconomic
variables that are considered in the literature to be important for explaining fiscal sustainability.
When examined in more detail, our results show that a reduction in government expenditures
for the two models – which is consistent with the improvement of public spending efficiency
scores – tends to have a positive impact on fiscal solvency. However, the results obtained for
Models 1 and 2 are not significant when applying the 2SLS estimator, whereby this estimator
presents a significant positive coefficient for efficient scores for Model 0, but only for cases of
correlation greater than 0.25 (see Appendix, Table A4).
Two major conclusions emerge from our results: the first is related to the fact that
improvements in public spending efficiency are more associated with higher fiscal
sustainability levels for the Euro-area countries, rather than for the remaining countries of our
sample. The second, and probably the most interesting, is the fact that rationalising public
expenditures, without jeopardising the actual level of public goods and the provision of services
is found to be a better determinant for fiscal sustainability than the improvement of the primary
budget balance. This result is obtained by comparing the magnitude of both coefficients when
both variables are statistically significant, which is a result that presents important policy
implications. In fact, this conclusion highlights other ways to lead public finances to a
sustainable path, beyond the traditional tax increase or spending cuts measures employed by
governments, usually without considering the efficiency of the public administration.
Additionally, we found that changes in government debt-to-GDP ratios have a non-
significant impact on fiscal solvency. However, the output gaps present a marginal and small
detrimental impact on fiscal sustainability, which is indeed a surprising result. Indeed, a positive
output gap would lead to inflationary pressures that would have the effect of increasing
government revenues, leading to more sustainable public finances. Furthermore, in the cases of
a positive gap, in theory, public authorities would need to increase taxes or reduce public
expenditures to correct the excessive demand for the existing supply. This can also be another

6
We report the results of Model 0 for both input-oriented and output-oriented approaches in the Appendix,.

10
way of improving fiscal sustainability. Therefore, and in order to explain such a result of the
output gap effect on fiscal sustainability, it would be beneficial to study how the elasticity of
government spending to GDP growth rate behaves during that same period, that is to say, to
assess whether a Wagner’s law event occurred.7 Lastly, although the interest rate-GDP growth
rate differentials are only significant for Model 0 (Table A4, in the Appendix), the negative
impact obtained for such a differential is to be expected. Indeed, the greater the costs derived
from debt interests when compared with the GDP growth effect, the greater the effect over
revenues, as well as the reduction of the debt-to-GDP ratio. Conversely, the capability of public
authorities in managing public debt in a sustainable path in the future becomes less.

7
For instance, Afonso and Alves (2017) assess Wagner’s law by function of the government. They found different
responses when an economy has a positive output gap when compared to the cases when a given economy is below
its potential GDP.

11
 Table 1. Estimations results for the impact of public spending efficiency on fiscal
sustainability, input-oriented scores, Model 1, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
(1) (2) (3)
𝛽𝑡−1 0.913*** 0.961*** 1.344*** 1.632*** 0.866*** 0.894***
(0.037) (0.038) (0.090) (0.424) (0.042) (0.052)
𝜃 0.058** 0.034 0.054 0.003 0.089*** 0.055**
(0.023) (0.022) (0.055) (0.197) (0.030) (0.025)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.002 0.003***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.002 0.000
(0.000) (0.005) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002** -0.010 -0.002*
(0.001) (0.025) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.003) (0.001)
Obs. 455 310 385 259 389 244
𝑅2 0.996 0.997 0.997 0.996 0.994 0.996
Correlation>0.25
OLS-FE 2SLS WLS-FE
(1) (2) (3)
𝛽𝑡−1 0.782*** 0.874*** 1.220*** 1.543*** 0.772*** 0.845***
(0.056) (0.055) (0.098) (0.255) (0.057) (0.058)
𝜃 0.084*** 0.014 0.003 0.008 0.106*** 0.054*
(0.029) (0.027) (0.040) (0.060) (0.030) (0.030)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.004*** 0.003* 0.003***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 0.000 -0.001 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.008 -0.002**
(0.001) (0.008) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.001) (0.001)
Obs. 169 116 143 97 148 95
𝑅2 0.996 0.997 0.997 0.997 0.995 0.996
Euro area
OLS-FE 2SLS WLS-FE
(1) (2) (3) (4) (5) (6)
𝛽𝑡−1 0.884*** 0.928*** 1.236*** 1.287*** 0.827*** 0.871***
(0.042) (0.041) (0.112) (0.196) (0.064) (0.061)
𝜃 0.152*** 0.079*** 0.044 -0.030 0.167*** 0.103***
(0.031) (0.025) (0.083) (0.192) (0.042) (0.028)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.003*** 0.003***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 0.000 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.004 -0.003***
(0.001) (0.005) (0.001)
𝑟−𝑔 -0.000 -0.001 -0.000
(0.000) (0.001) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.996 0.996 0.997 0.993 0.995
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
1 includes two inputs, governments’ normalised spending on opportunity and on “Musgravian” indicators, and one
output.

12
 Table 2. Estimations results for the impact of public spending efficiency on fiscal
sustainability, input-oriented scores, Model 2, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.911*** 0.957*** 1.337*** 1.618*** 0.909*** 0.908***
(0.037) (0.038) (0.088) (0.381) (0.034) (0.052)
𝜃 0.060** 0.037 0.056 -0.007 0.074*** 0.011
(0.023) (0.024) (0.049) (0.126) (0.023) (0.040)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.003 0.004***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.004) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002** -0.009 -0.002*
(0.001) (0.018) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.002) (0.001)
Obs. 455 310 385 259 455 244
𝑅2 0.996 0.997 0.997 0.996 0.997 0.996
Correlation>0.25
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.791*** 0.910*** 1.154*** 1.485*** 0.776*** 0.916***
(0.051) (0.038) (0.088) (0.160) (0.063) (0.047)
𝜃 0.102*** 0.060*** 0.042 0.084** 0.099*** 0.041
(0.025) (0.020) (0.032) (0.039) (0.029) (0.028)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.004*** 0.004***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002** -0.001 -0.002***
(0.001) (0.004) (0.001)
𝑟−𝑔 0.001 0.000 -0.000
(0.001) (0.001) (0.001)
Obs. 260 181 220 152 226 147
𝑅2 0.996 0.997 0.998 0.998 0.994 0.996
Euro Area
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.870*** 0.920*** 1.221*** 1.410*** 0.828*** 0.872***
(0.043) (0.041) (0.108) (0.364) (0.066) (0.061)
𝜃 0.141*** 0.080*** -0.026 0.133 0.143*** 0.089***
(0.032) (0.027) (0.090) (0.526) (0.041) (0.034)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.004*** 0.003***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 0.000 -0.000 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.003*** 0.001 -0.003***
(0.001) (0.009) (0.001)
𝑟−𝑔 -0.000 -0.001 -0.001
(0.001) (0.001) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.996 0.997 0.997 0.993 0.995
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
2 uses one input, governments’ normalised total expenditure and two outputs, the opportunity PSP and the
“Musgravian” PSP scores.

13
4.2.2. Output-Oriented Efficiency

Moving to the results of the effects of output-oriented efficiency scores on fiscal

sustainability, the linkage between higher efficiency generated by a higher provision of public
goods with the same level of inputs and fiscal sustainability is not as clear as it was for the
analysis of input-oriented related models. The rationale behind the results obtained for Models
1 and 2, as presented in Tables 3 and 4, respectively, is the following: for the same level of
inputs, higher efficiency would provide a larger provision of public goods and services. In the
same vein, the fiscal multiplier would be greater, causing more aggregate income and thus
generating more public revenues. This mechanism would increase government revenues and
also improve fiscal sustainability.
From our results for the output-oriented models, we found similar results to those
found for the input-oriented models. However, it is important to notice some important
differences: first, while we found higher effects of improving public administration efficiency
on fiscal sustainability for the whole sample and for Euro Area economies for the input-oriented
Model 1 when compared to the output-oriented approach for the same model, only the output-
oriented Model 1 provides better results for the efficiency scores effects over fiscal
sustainability. On the other hand, the results obtained for Model 2, in an output-oriented setup
(Table 4), always provide higher coefficients for public efficiency scores on fiscal
sustainability, whatever sample we chose to analyse.
When it comes to the control variables, we also reach similar conclusions to those found
for input-oriented models. Once again, while government debt-to-GDP increments are not
crucial to explain fiscal solvency, interest rate-GDP growth rate differential only matters when
analysing Model 0, and for the euro-area economies (see Table A5, in the Appendix).
Lastly, when comparing our overall set of results, there is a need to highlight the role of
public administration and how it is organised to improve and increases the provision of public
goods and services with the same level of public expenditure, both in an output-oriented
approach, or by providing the same amount of existing goods and services with less government
spending. In this context, we can come to the conclusion that programme budgeting could be
an important tool for improving such public spending efficiency. Furthermore, and in order to
provide a better understanding of these relationships, there is a need to analyse how inputs and
outputs impact fiscal sustainability to better design public administration rules and strategies.

14
 Table 3. Estimations results for the impact of public spending efficiency on fiscal
sustainability, output-oriented, Model 1, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.908*** 0.961*** 1.345*** 1.502*** 0.873*** 0.910***
(0.040) (0.039) (0.086) (0.155) (0.042) (0.052)
𝜃 0.025 0.017 0.016 -0.118 0.050** 0.018
(0.017) (0.017) (0.018) (0.463) (0.022) (0.026)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.002 0.004***
(0.001) (0.005) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.004) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.003*** -0.004 -0.002*
(0.001) (0.011) (0.001)
𝑟−𝑔 0.000 0.000 -0.000
(0.001) (0.003) (0.001)
Obs. 455 310 385 259 389 244
𝑅2 0.996 0.997 0.997 0.997 0.994 0.996
Correlation>0.25
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.692*** 0.890*** 1.030*** 0.374 0.634*** 0.843***
(0.087) (0.048) (0.110) (7.629) (0.092) (0.056)
𝜃 0.104*** 0.030 0.085*** 0.098 0.124*** 0.042
(0.033) (0.028) (0.033) (0.215) (0.036) (0.034)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.004 0.003***
(0.001) (0.004) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 0.001 0.000
(0.000) (0.013) (0.001)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 0.002 -0.002 0.001
(0.001) (0.044) (0.001)
𝑟−𝑔 0.001 -0.002 -0.000
(0.001) (0.013) (0.001)
Obs. 156 91 132 77 135 70
𝑅2 0.997 0.999 0.998 0.992 0.996 0.998
Euro Area
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.885*** 0.927*** 1.345*** 1.273*** 0.875*** 0.910***
(0.047) (0.041) (0.094) (0.203) (0.068) (0.061)
𝜃 0.050* 0.064** 0.066 -0.226 0.053 0.027
(0.030) (0.030) (0.061) (0.758) (0.036) (0.038)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.004*** -0.000 0.004***
(0.001) (0.010) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.005) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.004*** 0.004 -0.003**
(0.001) (0.027) (0.001)
𝑟−𝑔 -0.000 0.000 -0.000
(0.000) (0.005) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.996 0.996 0.993 0.992 0.994
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
1 includes two inputs, governments’ normalised spending on opportunity and on “Musgravian” indicators, and one
output.

15
 Table 4. Estimations results for the impact of public spending efficiency on fiscal
sustainability, output-oriented, Model 2, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.905*** 0.954*** 1.302*** 1.638*** 0.876*** 0.905***
(0.038) (0.038) (0.090) (0.288) (0.042) (0.052)
𝜃 0.102*** 0.109*** 0.006 0.060 0.118** 0.073
(0.034) (0.033) (0.051) (0.363) (0.050) (0.051)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002*** 0.003 0.004***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.003) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.009 -0.002*
(0.001) (0.016) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.002) (0.001)
Obs. 455 310 385 259 389 244
𝑅2 0.996 0.997 0.997 0.996 0.994 0.996
Correlation>0.25
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.750*** 0.954*** 1.060*** 1.638*** 0.769*** 0.905***
(0.070) (0.038) (0.134) (0.288) (0.087) (0.052)
𝜃 0.201*** 0.109*** -0.024 0.060 0.251*** 0.073
(0.061) (0.033) (0.086) (0.363) (0.082) (0.051)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002*** 0.003 0.004***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.003) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.009 -0.002*
(0.001) (0.016) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.002) (0.001)
Obs. 195 310 165 259 169 244
𝑅2 0.997 0.997 0.998 0.996 0.992 0.996
Euro Area
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.886*** 0.931*** 1.192*** 1.325*** 0.873*** 0.907***
(0.043) (0.039) (0.211) (0.201) (0.065) (0.061)
𝜃 0.182*** 0.119*** -0.158 -0.011 0.164** 0.069
(0.056) (0.039) (0.679) (0.791) (0.079) (0.061)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.004*** 0.003*** 0.004***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 0.000 -0.000 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.003 -0.003***
(0.001) (0.004) (0.001)
𝑟−𝑔 -0.000 -0.001 -0.000
(0.000) (0.002) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.997 0.996 0.997 0.992 0.994
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
2 uses one input, governments’ normalised total expenditure and two outputs, the opportunity PSP and the
“Musgravian” PSP scores.

16
5. Conclusion
In this paper we assessed to what extent better government spending efficiency
contributes to higher levels of fiscal sustainability, for a panel of 35 OECD countries during the
period of 2007-2020.
We first compute the magnitude of the response of government revenues to changes in
government spending, to test the hypothesis that both sides of the budget balance should move
together. Next, we make use of so-called government spending efficiency scores, which show
notably how governments could increase their performance whilst maintaining the same level
of inputs, or how governments can reduce the level of inputs, while maintaining the same level
of performance. Finally, we empirically evaluate the responsiveness of fiscal sustainability to
changes in government spending efficiency.
Regarding the answer to our research question, our results show notably that more
efficient governments contribute more to increased fiscal sustainability. For the case of the
input-oriented efficiency scores, the underlying rational implies that less public resources can
provide the same level of output and can directly improve the fiscal balance and fiscal
sustainability. In the case of the output-oriented efficiency scores, the explanation can be
explained by the provision of more and better government outputs, which affect higher
economic growth and greater government revenues, which in turn also improves fiscal
sustainability. More specifically, rationalising public expenditures without jeopardising the
actual level of public goods and the provision of services is found to be a better determinant for
fiscal sustainability than improving the primary budget balance.
In sum, on the one hand the policy implications of our overall set of results point to the
crucial role of the organisation of public administration for improving the provision of public
goods and services whilst maintaining the same level of public expenditure. On the other hand,
the same level of existing public goods and services could be guaranteed with less government
spending.

17
References
1. Adam, A., Delis, M., Kammas, P. (2011). Public sector efficiency: levelling the playing
field between OECD countries. Public Choice, 146(1-2), 163–183.
https://doi.org/10.1007/s11127-009-9588-7
2. Afonso, A. (2005). Fiscal Sustainability: The Unpleasant European Case.
FinanzArchiv/Public Finance Analysis, 61(1), 19-44. https://doi.org/10.1007/s11127-009-
9588-7
3. Afonso, A., Alves, J. (2017). Reconsidering Wagner’s law: evidence from the functions
of the government. Applied Economics Letters, 24(5), 346-350.
https://doi.org/10.1080/13504851.2016.1192267
4. Afonso, A., Alves, J., Jalles, J. (2022a). Fiscal Reaction Functions in the EU: a tale of
(r-g) and whether it matters. Révue française d´économie, XXXVI, 65-101.
https://doi.org/10.3917/rfe.213.0065
5. Afonso, A., Jalles, J., Venâncio, A. (2022b). Do Financial Markets Reward Government
Spending Efficiency? Journal of International Financial Markets, Institutions & Money,
77(101505), 1-21. https://doi.org/10.1016/j.intfin.2022.101505
6. Afonso, A., Jalles, J. (2017). Euro area time-varying fiscal sustainability. International
Journal of Finance and Economics, 22(3), 244-254. https://doi.org/10.1002/ijfe.1582
7. Afonso, A., Jalles, J., Venâncio, A. (2021). Taxation and Public Spending Efficiency:
An International Comparison. Comparative Economic Studies, 63(3), 356-383.
https://doi.org/10.1057/s41294-021-00147-2
8. Afonso, A., Kazemi, M. (2017). “Assessing Public Spending Efficiency in 20 OECD
Countries”, in Inequality and Finance in Macrodynamics (Dynamic Modeling and
Econometrics in Economics and Finance), Bökemeier, B., Greiner, A. (Eds), 23, Springer.
https://doi.org/10.1007/978-3-319-54690-2_2
9. Afonso, A., Rault, C. (2015). Multi-step analysis of public finances sustainability.
Economic Modelling, 48, 199-209. https://doi.org/10.1016/j.econmod.2014.10.011
10. Afonso, A., Romero, A., Monsalve, E. (2013). “Public Sector Efficiency: Evidence for
Latin America Public”. Inter-American Development Bank, 80478, Inter-American
Development Bank. http://dx.doi.org/10.2139/ssrn.2365007
11. Afonso, A., Schuknecht, L., Tanzi, V. (2005). Public Sector Efficiency: An
International Comparison. Public Choice, 123(3-4), 321-347. https://doi.org/10.1007/s11127-
005-7165-2

18
12. Afonso, A., Schuknecht, L., Tanzi, V. (2010). Public Sector Efficiency: Evidence for
New EU Member States and Emerging Markets. Applied Economics, 42(17), 2147-2164.
https://doi.org/10.1080/00036840701765460
13. Afonso, A., St. Aubyn, M. (2019). Economic Growth, Public, and Private Investment
Returns in 17 OECD Economies. Portuguese Economic Journal, 18(1), 47-65.
https://doi.org/10.1007/s10258-018-0143-7
14. Aldama, P. Creel, J. (2021). “Real-time fiscal policy responses in the OECD from 1997
to 2018: Procyclical but sustainable?” European Journal of Political Economy, 1-21.
https://doi.org/10.1016/j.ejpoleco.2021.102135
15. Amato, M., Saraceno, F. (2022). Squaring the Circle: How to Guarantee Fiscal Space
and Debt Sustainability with a European Debt Agency. BAFFI CAREFIN Centre Research
Paper No. 2022-172, 1-25. http://dx.doi.org/10.2139/ssrn.4011081
16. Antonelli, M., de Bonis, V. (2019). The efficiency of social public expenditure in
European countries: a two-stage analysis. Applied Economics, 51(1), 47-60.
https://doi.org/10.1080/00036846.2018.1489522
17. Bergman, U. M., Hutchison, M. M., Jensen, S. E. H. (2016). “Promoting sustainable
public finances in the European Union: The role of fiscal rules and government efficiency”.
European Journal of Political Economy, 44(1), 1-19.
https://doi.org/10.1016/j.ejpoleco.2016.04.005
18. Blanchard, O., Leandro, A., Zettelmeyer. J. (2020). Redesigning the EU Fiscal Rules:
From Rules to Standards. Mimeo, August 30, 2020. 72nd Economic Policy Panel Meeting 22-
23 October 2020, Supported by the Federal Ministry of Finance, Germany.
19. Bohn, H. (2008). “The Sustainability of Fiscal Policy in the United States”, in Neck, R.,
Sturm, J.-E. (eds.), Sustainability of Public Debt. The MIT Press, Cambridge, Massachusetts.
20. Brady, G., Magazzino, C. (2018), “Fiscal Sustainability in the EU”, Atlantic Economic
Journal, 46, 297-311.
21. Charnes, A.; Cooper, W., Rhodes, E. (1978). Measuring the efficiency of decision
making units. European Journal of Operational Research, 2(6), 429–444.
https://doi.org/10.1016/0377-2217(78)90138-8
22. Coelli T., Rao, D., Battese, G. (2002). An Introduction to Efficiency and Productivity
Analysis, 6th edition, Massachusetts, Kluwer Academic Publishers.
23. Dutu, R., Sicari, P. (2020). Public Spending Efficiency in the OECD: Benchmarking
Health Care, Education, and General Administration. Review of Economic Perspectives, 20(3),
253-280. https://doi.org/10.2478/revecp-2020-0013

19
24. Farrell, M. (1957). The Measurement of Productive Efficiency. Journal of the Royal
Statistical Society: Series A (General), 120, 253-281. https://doi.org/10.2307/2343100
25. Gupta, S., Verhoeven, M. (2001). The efficiency of government expenditure –
experiences from Africa. Journal of Policy Modelling, 23(4), 433-467.
https://doi.org/10.1016/S0161-8938(00)00036-3
26. Hakkio, C., Rush, M. (1991). Is the Budget Deficit “Too Large?”. Economic Inquiry,
29(3), 429-445. https://doi.org/10.1111/j.1465-7295.1991.tb00837.x
27. Hamilton, J., Flavin, M. (1986). On the Limitations of Government Borrowing: A
Framework for Empirical Testing. The American Economic Review, 76(4), 808-819.
28. Hauner, D. Kyobe, A. (2010). Determinants of Government Efficiency. World
Development, 38(11), 1527-1542. https://doi.org/10.1016/j.worlddev.2010.04.004
29. Leeper, E. (1991). Equilibria under 'Active' and 'Passive' Monetary and Fiscal Policies.
Journal of Monetary Economics, 27(1), 129-147. https://doi.org/10.1016/0304-
3932(91)90007-B
30. Magazzino, C., Brady, G., Forte, F. (2019). A panel data analysis of the fiscal
sustainability of G-7 countries. The Journal of Economic Asymmetries, 20, 1-9.
https://doi.org/10.1016/j.jeca.2019.e00127
31. Mauro, P., Romeu, R., Binder, A., Zaman, A. (2013). “A Modern History of Fiscal
Prudence and Profligacy”. IMF Working Paper No. 13/5, International Monetary Fund,
Washington, DC.
32. Mohanty, R., Bhanumurthy, N., Sahoo, B. (2022). “Analysing Public Sector Efficiency
of the Indian States”, in Afonso, A., Jalles, J., Venâncio, A. (eds.) Handbook on Public Sector
Efficiency, Edward Elgar Publishing, forthcoming.
33. Olanubi, S., Olanubi, O. (2022). “Political Short-termism and Government Spending
Efficiency in Sub-Sahara Africa”, in Afonso, A., Jalles, J., Venâncio, A. (eds.) Handbook on
Public Sector Efficiency, Edward Elgar Publishing, forthcoming.
34. Papadopoulos, A., Sidiropoulos, M. (1999). The sustainability of fiscal policies in the
European Union. International Advances in Economic Research, 5(3), 289-307.
https://doi.org/10.1007/BF02296413
35. Sims, C. (1994). A Simple Model for the Study on the Determination of the Price Level
and the Interaction of Monetary and Fiscal Policy. Economic Theory, 4(3), 381-399.
https://doi.org/10.1007/BF01215378

20
36. Trehan, B., Walsh, C. (1991). Testing Intertemporal Budget Constraints: Theory and
Applications to U. S. Federal Budget and Current Account Deficits. Journal of Money, Credit
and Banking, 23(2), 206-223. https://doi.org/10.2307/1992777
37. Vanhorebeek, F., Van Rompuy, P. (1995). Solvency and sustainability of fiscal policies
in the EU. De Economist, 143(4), 457-473. https://doi.org/10.1007/BF01384910
38. Woodford, M. (1995). Price Level Determinacy without Control of Monetary
Aggregate. Carnegie-Rochester Conference Series on Public Policy, 43, 1-46.
https://doi.org/10.1016/0167-2231(95)90033-0

21
Appendix

Table A1. Data series and sources.

Variable Series Source
Input, output efficiency scores
ltein0, ltein1, ltein2, teout0n, teout1n, teout2n Afonso et al. (2022b)
Government spending General government total expenditure WEO and Mauro et al. (2013)
Government revenue General government revenue WEO and Mauro et al. (2013)
Government debt General government gross debt WEO
Output gap Output gap in percent of potential GDP WEO
The series was computed by the ratio between
government spending on government debt's
interests and the government debt, both in
Interest rate GDP terms; The expenditures on interest were Own calculations
obtained by calculating the difference between
the primary and global budget balances, both
with series from the WEO
GDP growth rate Annual GDP growth rate WEO
Notes: WEO – World Economic Outlook, October 2021.

22
 Table A2. Augmented Dickey-Fuller and Phillips-Perron Unit Root tests for revenues and expenditures (% of GDP), 1980-2020.
Revenues Expenditures
ADF PP ADF PP
Levels Obs. F.D. Obs. Levels Obs. F.D. Obs. Levels Obs. F.D. Obs. Levels Obs. F.D. Obs.
Australia -2.610 39 -3.573** 38 -2.274 40 -5.934*** 39 -1.617 39 -2.498*** 38 -1.172 40 -2.557 39
Austria -2.222 39 -5.263*** 38 -2.387 40 -6.097*** 39 -2.851* 39 -3.528** 38 -2.766* 40 -3.989*** 39
Belgium -1.803 39 -4.105*** 38 -1.866 40 -5.715*** 39 -1.961 39 -2.938* 38 -1.663 40 -6.841*** 39
Canada -1.654 39 -3.552** 38 -2.016 40 -4.454*** 39 -1.791 39 -2.443 38 -1.595 40 -2.605* 39
Chile -2.767* 39 -5.088*** 38 -2.626* 40 -6.038*** 39 -2.11 39 -3.803*** 38 -1.524 40 -4.088*** 39
Colombia -1.062 39 -4.721*** 38 -0.974 40 -5.743*** 39 -0.397 39 -4.506*** 38 -0.384 40 -6.391*** 39
Czech Republic -1.723 24 -4.109*** 23 -2.256 25 -6.091*** 24 -2.727* 24 -3.551** 23 -5.391*** 25 -6.613*** 24
Denmark -2.939* 39 -4.583*** 38 -2.596* 40 -5.344*** 39 -3.046** 39 -4.357*** 38 -2.629* 40 -4.979*** 39
Finland -2.559 39 -3.506** 38 -3.264** 40 -7.108*** 39 -2.777* 39 -3.844*** 38 -2.138 40 -3.461*** 39
France -1.527 39 -3.786*** 38 -1.571 40 -5.338*** 39 -0.716 39 -3.683*** 38 -1.388 40 -4.004*** 39
Germany -1.689 39 -4.851*** 38 -1.973 40 -7.05*** 39 -2.795* 39 -4.708*** 38 -3.109** 40 -6.887*** 39
Greece -1.144 39 -4.556*** 38 -0.784 40 -5.655*** 39 -1.464 39 -3.954*** 38 -1.613 40 -5.352*** 39
Hungary -2.23 24 -3.731** 23 -2.057 25 -4.4*** 24 -3.418** 24 -4.088*** 23 -4.34*** 25 -4.473*** 24
Iceland -2.192 39 -5.455*** 38 -3.079** 40 -10.191*** 39 -2.116 39 -4.827*** 38 -2.224 40 -6.314*** 39
Ireland 0.776 39 -4.608*** 38 0.741 40 -6.238*** 39 -1.644 39 -4.273*** 38 -1.623 40 -6.286*** 39
Israel -1.550 39 -4.559*** 38 -1.681 40 -6.507*** 39 -1.986 35 -4.294*** 33 -2.191 37 -6.949*** 35
Italy -2.911 39 -4.759*** 38 -2.525 40 -5.361*** 39 -1.954 39 -2.605 38 -2.777* 40 -3.424** 39
Japan -0.797 39 -4.036*** 38 -0.623 40 -5.752*** 39 -0.401 39 -2.817* 38 -0.315 40 -3.363** 39
Latvia -0.859 21 -2.906* 20 -0.845 22 -3.518*** 21 -2.282 21 -3.811*** 20 -1.783 22 -3.277** 21
Lithuania -3.058** 24 -4.256*** 23 -1.727 25 -5.146*** 24 -2.513 24 -2.191 23 -2.129 25 -3.353** 24
Luxembourg -2.794* 24 -4.998*** 23 -2.613* 25 -4.693*** 24 -3.826*** 24 -4.324*** 23 -2.604* 25 -3.281** 24
Netherlands -1.203 39 -4.534*** 38 -1.108 40 -5.847*** 39 -1.546 39 -3.953*** 38 -1.414 40 -5.859*** 39
New Zealand -1.565 34 -2.663* 33 -1.019 35 -3.703*** 34 -2.479 34 -3.500** 33 -1.603 35 -2.799* 34
Norway -2.429 39 -5.077*** 38 -2.35 40 -5.607*** 39 -1.845 39 -4.610*** 38 -1.517 40 -4.133*** 39
Poland -5.318*** 24 -4.614*** 23 -2.613* 25 -5.758*** 24 -3.860*** 24 -1.761 23 -2.346 25 -3.918*** 24
Portugal -1.964 33 -7.387*** 32 -2.301 34 -6.923*** 33 -3.109** 33 -4.93*** 32 -3.183** 34 -4.039*** 33
Slovakia -1.637 24 -2.460 23 -2.004 25 -5.757*** 24 -2.552 24 -4.049*** 23 -2.060 25 -5.481*** 24
Slovenia -2.246 24 -3.333** 23 -1.957 25 -5.224*** 24 -2.371 24 -3.508** 23 -3.294** 25 -6.222*** 24
South Korea -0.924 39 -3.824*** 38 -0.949 40 -6.334*** 39 -0.452 39 -3.477** 38 -0.488 40 -7.087*** 39
Spain -2.265 39 -7.303*** 38 -1.304 40 -9.891*** 39 -1.863 39 -5.595*** 38 -1.104 40 -9.122*** 39
Sweden -1.914 39 -4.806*** 38 -2.14 40 -6.282*** 39 -2.344 39 -4.024*** 38 -2.301 40 -5.301*** 39
Switzerland -1.915 39 -3.263** 38 -1.958 40 -6.211*** 39 -2.024 39 -3.245** 38 -1.734 40 -3.643*** 39
Turkey -0.855 39 -3.078** 38 -0.606 40 -4.041*** 39 -1.178 39 -3.55** 38 -1.071 40 -5.307*** 39
UK -1.812 39 -3.897*** 38 -1.309 40 -6.299*** 39 -3.156** 39 -2.522 38 -2.137 40 -3.085** 39
US -2.047 39 -4.82*** 38 -1.937 40 -5.577*** 39 -2.480 39 -2.257*** 38 -2.048 40 -2.250 39

23
Table A3. Engle-Granger cointegration test results.
Z(t)
Australia -3.652**
Austria -5.425***
Belgium -3.994***
Canada -3.649**
Chile -4.772***
Colombia -5.006***
Czech Republic -4.355**
Denmark -4.608***
Finland -4.255***
France -3.686**
Germany -4.906***
Greece -4.406***
Hungary -3.983**
Iceland -5.106***
Ireland -4.669***
Israel -3.375*
Italy -4.919***
Japan -4.107**
Latvia -3.925**
Lithuania -4.742***
Luxembourg -4.527***
Netherlands -4.540***
New Zealand -3.708**
Norway -4.262**
Poland -5.063***
Portugal -8.086***
Slovakia -6.779***
Slovenia -3.414*
South Korea -4.643***
Spain -4.963***
Sweden -3.524**
Switzerland -3.170*
Turkey -3.228*
UK -3.873**
US -4.903***

24
 Table A4. Estimations results for the impact of public spending efficiency on fiscal sustainability, input-
oriented, Model 0, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.914*** 0.961*** 1.349*** 1.592*** 0.870*** 0.896***
(0.037) (0.038) (0.089) (0.348) (0.042) (0.053)
𝜃 0.054** 0.035 0.057 -0.020 0.088*** 0.059**
(0.024) (0.025) (0.055) (0.235) (0.028) (0.025)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.003* 0.003***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 0.000
(0.000) (0.004) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002** -0.008 -0.002**
(0.001) (0.018) (0.001)
𝑟−𝑔 0.000 -0.000 -0.000
(0.001) (0.002) (0.001)
Obs. 455 310 385 259 389 244
𝑅2 0.996 0.997 0.997 0.997 0.994 0.996
Correlation>0.25
OLS-FE WLS-FE
2SLS
𝛽𝑡−1 0.765*** 0.907*** 1.062*** 1.486*** 0.727*** 1.486***
(0.055) (0.036) (0.094) (0.159) (0.067) (0.159)
𝜃 0.096*** 0.034 0.025 0.132* 0.112*** 0.132*
(0.032) (0.021) (0.043) (0.074) (0.032) (0.074)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.004*** 0.004***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.001 -0.001
(0.000) (0.001) (0.001)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.001 0.004 0.004
(0.001) (0.004) (0.004)
𝑟−𝑔 0.000 -0.001 -0.001
(0.000) (0.001) (0.001)
Obs. 195 142 165 119 161 119
𝑅2 0.996 0.998 0.998 0.999 0.995 0.999
Euro Area
OLS-FE WLS-FE
2SLS
𝛽𝑡−1 0.880*** 0.926*** 1.242*** 1.351*** 0.846*** 0.884***
(0.042) (0.041) (0.107) (0.143) (0.064) (0.061)
𝜃 0.141*** 0.082*** 0.015 0.051 0.142*** 0.094***
(0.033) (0.024) (0.072) (0.142) (0.040) (0.028)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.004*** 0.003***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 0.000 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.003*** -0.001 -0.003***
(0.001) (0.004) (0.001)
𝑟−𝑔 -0.000 -0.001** -0.000
(0.000) (0.000) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.997 0.996 0.998 0.993 0.995
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
0 includes only one input (government spending as percentage of GDP) and one output, a composite public sector
performance (PSP) indicator.

25
 Table A5. Estimations results for the impact of public spending efficiency on fiscal sustainability, output-
oriented, Model 0, 2007-2020.
Baseline
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.908*** 0.961*** 1.347*** 1.494*** 0.874*** 0.911***
(0.040) (0.039) (0.088) (0.139) (0.043) (0.052)
𝜃 0.023 0.015 0.005 0.027 0.040* 0.013
(0.017) (0.018) (0.017) (0.087) (0.020) (0.027)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.003*** 0.004*** 0.004***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 0.000 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.002*** -0.001 -0.002*
(0.001) (0.003) (0.001)
𝑟−𝑔 0.000 -0.001 -0.000
(0.001) (0.001) (0.001)
Obs. 455 310 385 259 389 244
𝑅2 0.996 0.997 0.997 0.998 0.994 0.996
Correlation>0.25
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.725*** 0.888*** 1.044*** 1.280*** 0.701*** 0.844***
(0.087) (0.048) (0.113) (0.397) (0.100) (0.054)
𝜃 0.091*** 0.035 0.055* 0.140** 0.101*** 0.044
(0.030) (0.031) (0.028) (0.062) (0.031) (0.036)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.002** 0.003*** 0.003***
(0.001) (0.001) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 -0.000 0.000
(0.000) (0.000) (0.001)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 0.002 0.003 0.001
(0.001) (0.002) (0.002)
𝑟−𝑔 0.001 -0.000 -0.000
(0.001) (0.001) (0.001)
Obs. 156 91 132 77 135 70
𝑅2 0.996 0.999 0.998 0.999 0.994 0.998
Euro Area
OLS-FE 2SLS WLS-FE
𝛽𝑡−1 0.888*** 0.928*** 1.361*** 1.191*** 0.879*** 0.912***
(0.047) (0.040) (0.091) (0.217) (0.067) (0.061)
𝜃 0.046 0.065** 0.067 0.111 0.046 0.021
(0.031) (0.032) (0.064) (0.277) (0.038) (0.040)
𝑝𝑏𝑎𝑙𝑎𝑛𝑐𝑒 0.004*** 0.004** 0.004***
(0.001) (0.002) (0.001)
Δ𝑑𝑒𝑏𝑡 -0.000 0.001 0.000
(0.000) (0.001) (0.000)
𝑜𝑢𝑡𝑝𝑢𝑡𝑔𝑎𝑝 -0.004*** -0.004 -0.003**
(0.001) (0.005) (0.001)
𝑟−𝑔 -0.000 -0.001* -0.000
(0.000) (0.001) (0.001)
Obs. 194 179 167 149 181 166
𝑅2 0.995 0.996 0.996 0.997 0.991 0.994
Note: Constant term, country and time effects estimated and omitted for reasons of parsimony. Robust standard
errors in parenthesis. *, **, *** denote statistical significance at the 10, 5, and 1 percent levels, respectively. Model
0 includes only one input (government spending as percentage of GDP) and one output, a composite public sector
performance (PSP) indicator.

26
 Table A6. Correlation Matrix of Efficiency Scores for Input-Oriented Models.
ltein0 ltein1 ltein2
ltein0 1.000
ltein1 0.955 1.000
ltein2 0.951 0.921 1.000

Table A7. Correlation Matrix of Efficiency Scores for Output-Oriented Models.

teout0n teout1n teout2n
teout0n 1.000
teout1n 0.982 1.000
teout2n 0.688 0.691 1.000

Model 0, one input, governments’ normalised spending, and one output, total PSP scores (ltein0; teout0n).
Model 1, two inputs and one output (ltein1, teout1n). Model 2, one input and two outputs (ltein2, teout2n).

27