by Pierre
Aldama

Between 1999 and 2019, the eve of the Covid-19
pandemic, the public debts of the 11 oldest euro zone members had risen by
an average of 20 percentage points of GDP. This increase in public debt is
commonly attributed to structural budget deficits, particularly those in the
pre-crisis period and in the “South”. But how much of the stock of public debt
in 2019 can be attributed to structural deficits, and how much to GDP growth,
interest payments or cyclical deficits? In this post, we use the December 2020
edition of the OECD’s Economic
Outlook to break down the changes in public debt into its main factors:
structural and cyclical primary balances, the interest burden, nominal GDP
growth and stock-flow adjustments. This shows that the structural deficits
generally contributed less than is commonly assumed, and that the increase in
public debt over the period was largely the result of the direct and indirect
consequences of the double-dip recession in the euro zone.

On the eve of the Covid-19 crisis, the 11 oldest
euro zone countries had an average level of public debt (in the Maastricht
sense) of 92% of GDP. Between 1999 and 2019, the public debt in these 11
countries increased by an average of 20 percentage points of GDP, although with
considerable heterogeneity (Figure 1). On the one hand, a group of so-called
virtuous countries – Germany, the Netherlands, Austria, Finland and Ireland – reduced
their debt ratios to their 1999 level of 60% of GDP or even lower. In contrast
to this were the countries whose public debt increased – France, Spain, Greece
and Portugal – or remained at a high level – Belgium and Italy. Can we simply
deduce from this that there are some countries that acted like the proverbial
ant and others like the grasshopper? Probably not.

Indeed, not all countries entered the European
Monetary Union (EMU) with the same level of debt: their starting point
therefore biases observation insofar as it does not inform about the structural
or cyclical factors or to the interest burden associated with the fiscal policy
in place from 1999 to 2019. Is the rise in public debt in the “grasshopper” countries
largely attributable to the accumulation of structural deficits, or on the
contrary, to cyclical factors and the impact of the recessions in the euro zone
(2008-2010 and 2011-2013)?

This post uses the December 2020 edition of the
OECD’s Economic Outlook to break down the changes inpublic debt into the main components: structural
and cyclical primary balances, the interest burden, nominal GDP growth and
stock-flow adjustments. This shows that the contribution of structural deficits
is generally lower than commonly assumed and that the increase in public debt
over the period largely results from the direct and indirect consequences of
the double-dip recession in the euro zone.

The accounting decomposition of public debt
dynamics

The change in public debt (as a percentage of GDP)
between year t and year t-1 can be broken
down into five main factors, using the following equation:

where r_t / (1+y_t) d_t-1 is
the effect of the interest burden, –y_t / (1+y_t)d_t-1 is
the effect of nominal GDP growth (and the sum of the two terms is the infamous
snowball effect[1] of public debt), sp_t^cyc is
the cyclical component of the primary budget balance (excluding the interest
burden), sp_t^struc is
the structural primary balance (adjusted for the output gap) and afs_t represents
the stock-flow adjustments, i.e. transactions on the assets and liabilities of
general government that are not accounted for in the primary balance.

By aggregating each of these terms, we calculate
the contributions to the total change in public debt between 1999 and 2019
(Figure 2) and year by year (Figure 3). Finally, Figures 4A and 4B present breakdowns
of the public debt similar to Figure 2 but over two sub-periods: 1999-2008 and
2008-2019.

The scars of the double recession of 2008-2010 and
2011-2013 in the euro zone

The rise in public debt in the EMU is largely
explained by the cyclical effects of the double recession of 2008-2010 and
2011-2013 (Figure 3). Between 2008 and 2019, in the three countries with the
largest increases in public debt (Greece, Spain, Portugal), the rise in debt is
due largely to cyclical primary deficits and the snowball effect. Greece is a
striking example: the snowball effect accounts for almost 3/5 of the increase
in public debt between 1999 and 2019, and this is concentrated mainly between
2008 and 2019, with the collapse of the level of GDP. In contrast, the apparent
Irish “miracle” is actually due to massive nominal growth in 2015, which in
turn is explained by the relocation of existing intangible assets in
Ireland by multinationals.

Moreover, any positive contribution of structural deficits to debt growth during the 2008-2010
crisis is in fact an optimal countercyclical response of fiscal policy during
the recession, and cannot be interpreted as a lack of fiscal seriousness per
se. This was the case, however, in fewer than half of the countries
studied: Spain, the Netherlands, France, Austria, and Ireland, and for the
other countries this largely reflects the pro-cyclical character of
discretionary fiscal policies in the euro zone over the period (Aldama and Creel, 2020).

Finally, in general, the contribution of the stock-flow
adjustments increases sharply after the 2008 crisis, mainly due to the banking
sector rescue plan. In the case of Greece, the negative contribution of these
adjustments largely corresponds to the 2012 default.

Northern surpluses vs. Southernstructural
deficits in the euro zone?

Over the period 1999-2019, it appears that only
three countries (France, Ireland and Portugal) showed a positive contribution
of structural primary deficits to the rise in public debt. Remarkably, both
Greece and Italy stand out from these countries with a negative contribution
due to their structural primary surpluses, as shall be seen later, due in
particular to the structural fiscal adjustment carried out since 2010 in the
case of Greece. Belgium, which was heavily indebted at the time of its entry
into the EMU (114% of GDP), is also characterised by the strong negative
contribution of its structural primary balance to debt growth.

In the case of Greece, we observe in particular the
sharp decline in the contribution of the structural primary balance, which even
becomes negative in 2019: in other words, by 2010 Greece has more than offset
the effect of its previous structural primary deficits. Even more remarkably,
Italy has pursued a very tight fiscal policy over the entire period, in so far as the (negative) contribution
of its structural primary surplus has steadily increased in absolute terms.
Portugal lies in between, and started to run structural primary surpluses,
without cancelling out the effect of its pre-2010 deficits. Ireland, sometimes
presented as the “good pupil” in the euro area following the 2010
crisis, did not have post-crisis structural surpluses that offset the
structural deficits run up during the crisis (the contribution to the change in
debt was stable).

Focusing on the pre-2008 period (Figure 4A) and the
so-called Southern countries, again only Greece and Portugal saw a positive
contribution of their structural deficits to debt growth, while the
contribution of the primary structural surpluses in Ireland, Italy and Spain was
negative.

On the Franco-German side, the divergence is clear.
German fiscal rigour appears almost extreme: even following the 2008-2010
crisis, the federal government’s primary structural balance did not contribute
positively to debt growth, reflecting a very weak countercyclical discretionary
policy (the German structural balance increased by 1 GDP point in 2010).
Conversely, in the case of France, a large part of the variation in public debt
can be explained by the structural deficits recorded both
before and after 2008 (Figures 4A and 4B), although this slowed down
in the second half of the 2010s (Figure 3). Thus, of the 37 GDP points of
public debt accumulated since 1999, almost 26 points came from structural
deficits accumulated over the period.

Of course, the distinction between the structural balance
and the cyclical balance is critically based on the estimation of the level of
“potential” GDP, i.e. of full utilization of production factors,
without inflationary pressures. This measure is subject to great uncertainty,
and there have been many criticisms, such as that it is too sensitive to the
macroeconomic cycle and to demand shocks (Coibion et al. 2018; Fatas and Summers 2018). Some studies suggest that the level of potential
activity may be underestimated. This likely bias in potential GDP estimates points
to the need for a note of caution about any definitive interpretation of the
structural vs. cyclical nature of budget deficits or surpluses. [2]

***

While public debt has increased overall in the euro
zone since 1999, a large part of this growth is explained by the direct and
indirect consequences of the 2008 crisis, through cyclical deficits, the
aggravation of the snowball effect and the structural weakness
of growth in certain Southern European countries.

On the contrary, most of the more indebted
countries today ran high primary structural surpluses over the period, such as
Italy and Belgium. Greece has even more than offset the positive contribution
of its past structural deficits. This is the reason why a reading grid that is
still overly used, that of the North versus the South, or of fiscal strictness versus
fiscal leniency, cannot stand up to a simple accounting analysis of the
dynamics of public debt.

---

[1] The snowball effect of public debt is the effect of the differential between the interest rate paid on the accumulated stock of debt and the economy’s growth rate. If this differential is positive, then for a given primary budget balance public debt tends to increase mechanically; conversely, if it is negative, public debt tends to decrease mechanically.

2] However, using the OECD Economic Outlook
has the advantage of providing a homogeneous approach across countries, and
therefore a relatively uniform bias between them. Moreover, the measure of
potential GDP used by the OECD is less cyclical than the measures used by the IMF and
the European Commission.

By Christophe Blot and Paul Hubert

In response to the health and economic crisis,
governments have implemented numerous emergency measures that have pushed public
debt up steeply. They have nevertheless not experienced any real difficulty in
financing these massive new issues: despite record levels of public debt, the
cost has fallen sharply (see Plus ou moins de
dette publique en France ?, by Xavier
Ragot). This trend is the result of
structural factors related to an abundance of savings globally and to strong
demand for secure liquid assets, characteristics that are generally met by
government securities. The trend is also related to the securities purchasing programmes
of the central banks, which have been stepped up since the outbreak of the
pandemic. For the year 2020 as a whole, the European Central Bank acquired
nearly 800 billion euros worth of securities issued by the governments of the
euro zone countries. In these circumstances, the central banks are holding an
increasingly high fraction of the debt stock, leading to a de facto
coordination of monetary and fiscal policies.

Back in 2009, central banks launched asset purchase
programmes to reinforce the expansionary impact of monetary policy in a context
where the banks’ key interest rates had reached a level close to 0%[1]. The stated objective was mainly to ease financing
conditions by holding down long-term interest rates on the markets. This
resulted in a sharp increase in the size of the banks’ balance sheets, which
now represents more than 53 GDP points in the euro zone and 35 points in
the United States, with the record being held by the Bank of Japan, at 133 GDP points
(Figure 1). These programmes, financed by issuing reserves, have focused heavily on government securities,
meaning that a large proportion of the stock of government debt is now held by
central banks (Figure 2). This proportion reaches 43% in Japan, 22% in the
United States and 25% in the euro zone. In the euro zone, in the absence of
euro bonds, the distribution of securities purchases depends on the share of
each national central bank in the ECB’s capital. The ECB’s distribution key stipulates
that the purchases are to be made pro rata to the share of the ECB’s capital
held by the national central banks[2]. Consequently, the purchases of securities are
independent of the levels and trajectories of public debt. As the latter are
heterogeneous, there are differences in the share of public debt held by the
national central banks [3]. Thus, 31% of Germany’s public debt is held by the
Eurosystem compared to 20% of Italy’s public debt.

The decentralization of fiscal policies in the euro zone is also leading to tensions in the sovereign debt markets of some member countries, as seen between 2010 and 2012 and more recently in March 2020. This is why Christine Lagarde has launched a new asset purchase programme called the Pandemic emergency purchase programme (PEPP). While the distribution key is not formally abolished, it may be applied more flexibly in order to allow the ECB to reduce the sovereign spreads between member countries. Analysing the flows of securities purchases made by the euro zone central banks and the debt issues of the member states, it can be seen that the Eurosystem has absorbed on average 72% of the public debt issued in 2020, i.e. 830 billion euros out of the 1155 billion of additional public debt. The share amounts to 76% for Spain, 73% for France, 70% for Italy and 66% for Germany (Figure 3).

Unlike purchases made under the APP programme,
which aim to hit the inflation target, the PEPP’s objective is first and
foremost to limit rate spreads, as Christine Lagarde reminded us on 16 July 2020.
In fact, even if there is a structural downward trend in interest rates, some
markets may be exposed to pressure. The euro zone countries are all the more
exposed as investors can arbitrate between the different markets without incurring
any exchange rate risks. This is why they may prefer German securities to
Italian securities, thereby undermining the homogeneous transmission of
monetary policy within the euro zone. In addition to arguments about the risk
of fragmentation, these operations also reflect a form of implicit coordination
between the single monetary policy and fiscal policies, providing countries
with the manoeuvring room needed to take the measures required to deal with the
health and economic crisis. By declaring on 10 December that the allocation
to the programme would increase to 1850 billion euros by no later than March
2022, the ECB sent a signal that it would maintain its support throughout the
duration of the pandemic[4].

---

[1] This policy, generally referred to as
quantitative easing (QE), was launched in March 2009 by the Bank of England and
the US Federal Reserve. Japan had already initiated this type of so-called
unconventional measure between 2001 and 2006, and resumed this approach in
October 2010. As for the ECB, the first purchases of securities targeted at
certain countries in crisis were made from May 2010. But it was not until March
2015 that a QE programme comparable to those implemented by the other major
central banks was developed.

[2] In practice, this share is relatively close
to the weight of each member country’s GDP in euro zone GDP.

[3] Securities purchasing operations are
decentralized at the level of the national central banks. Doing this reduces
risk-sharing within the Eurosystem since any losses would be borne by the
national central banks, unlike assets held directly by the ECB, for which there
is risk-sharing that depends on the share of each national central bank in the
ECB’s capital.

[4] The initial allocation was 750 billion euros,
which was increased in June 2020 by a further 600 billion. As of 31 December 2020,
securities purchases under the PEPP came to 650 billion.