Download Main results of the March 2004 fiscal notifications by the Candidate

EUROPEAN
ECONOMY
EUROPEAN COMMISSION
DIRECTORATE-GENERAL FOR ECONOMIC
AND FINANCIAL AFFAIRS
ENLARGEMENT PAPERS
ISSN 1608-9022
http://europa.eu.int/comm/economy_finance
N° 22 - December 2004
Main results of the March 2004 fiscal
notifications by the Candidate Countries
by
Directorate General for Economic
and Financial Affairs
KC-AA-04-022-EN-C
ISBN 92-894-5974-3
ECFIN/D/I/REP/50577 – EN
© European Communities, 2004.
Table of contents
page
1. Background.................................................................................................................... 3
2. Reported general government balances and debt levels................................................ 3
3. Economic background ................................................................................................... 4
4. Quality and reliability of the notified deficit and debt figures ...................................... 5
Evaluation of the fiscal notification of Bulgaria ................................................................ 6
Evaluation of the fiscal notification of Romania.............................................................. 12
Evaluation of the fiscal notification of Turkey................................................................. 19
MAIN RESULTS OF THE MARCH 2004 FISCAL NOTIFICATIONS
PRESENTED BY THE CANDIDATE COUNTRIES
1. BACKGROUND
All candidate countries1 were to submit fiscal notifications to the Commission services by
1 March 20042, in accordance with the commitments taken in the framework of the preaccession fiscal surveillance procedure. This notification exercise is the fourth produced by
the candidate countries after the 2001, 2002 and 2003 exercises. Countries generally met
this reporting deadline.
The framework of the fiscal notifications is now well known. All countries are familiar
with the EU legal and methodological principles for the calculation of general government
deficits and debt levels. The presentation of the notification tables and in particular the
reconciliation between the national budget balance and the balance used in EU fiscal
surveillance are both well understood. Significant efforts have been made to provide figures
that comply as much as possible with the methodology and coverage required by the fiscal
notifications.
The perspective of accession is a catalyst for reforms in the scope and the management of
the national budgets. Budget presentations are being modernised, and the coverage of
government operations by the national budgets is made more exhaustive. In particular,
there has been a spectacular reduction in the number of off-budget and special funds
accounts and operations. Also, the reference to central and general government in EU fiscal
surveillance often leads to more systematic monitoring, supervision and controllability of
the operations of local authorities and of social security.
2. REPORTED GENERAL GOVERNMENT BALANCES AND DEBT LEVELS
Table 1 shows the general government net borrowing/net lending figures reported in March
2004 and compares them with the figures reported in April 2003.
Table 1: General government net lending (+) / borrowing(-) (% of GDP)
Notification
Bulgaria
Romania
Turkey
2004
2003
2004
2003
2004
2003
1999
0.4
-4.5
-18.4
2000
2001
2002
-0.5
-0.5
-4.4
-4.6
-6.1
-5.8
0.2
0.2
-3.5
-3.3
-29.8
-26.9
-0.8
-0.6
-2.0
-2.2
-12.7
-10.0
(1)
2003 2004
-0.1
-0.7
-2.0
-2.5
-8.7
-11.3
average
1999-2003
-0.7
-0.2
-3.0
-3.3
-8.0
-15.1
(1) planned
1
Croatia has been a candidate country since June 2004 and will therefore start to participate in the preaccession reporting of deficits and debt as of 2005.
2
In 2004, the requested submission date was changed from 1 April, as in previous years, to 1 March, in
order to align the date to that of Member States’ spring notifications and in order to receive the
notifications in time for the preparation of the Commission Spring economic forecasts.
-3-
The 2004 notifications show that the general government deficits generally shrank in 2003
as compared to 2002, except for Romania, where it was unchanged. The plans for the
current year foresaw in the 2004 notifications a widening of the deficits in the case of
Bulgaria and Romania, and a further reduction in the case of Turkey3. Looking at trends
over a longer 5-year period, Bulgaria’s government balance was overall close to zero,
Romania’s deficit showed a slight improvement, mostly due to falling interest payments,
whereas Turkey’s deficits widened sharply during the 1999 and 2001 economic and
financial crises in the country but have, since then, come down considerably due to high
primary surpluses and falling interest rates.
Table 2 displays the general government gross debt ratios notified in March 2004 and
compares them with the figures submitted in April 2003.
Table 2: General government gross debt (% of GDP)
Notification
Bulgaria
Romania
Turkey
2004
2003
2004
2003
2004
2003
1999
79.3
24.0
67.4
2000
2001
2002
(1)
2003 2004
73.6
73.6
23.9
23.9
57.4
57.6
66.2
66.4
23.2
23.1
105.2
105.4
53.2
53.0
23.3
22.7
94.9
95.0
46.2
52.0
21.8
23.7
87.1
86.3
average
1999-2003
45.1
63.7
23.5
23.2
85.0
82.4
(1) planned
The gross debt ratios for the three candidate countries continued to vary considerably
between 22% for Romania and 87% for Turkey at the end of 2003. However, there has
been a continued trend of falling general government debt ratios for Bulgaria and Turkey,
which had exhibited high debt levels still some years ago, and a stabilisation of the
comparably low level for
Romania.
In Bulgaria and
Table 3: Main economic trends
Turkey, positive primary balances
annual
1999
2000
2001
2002
2003
averages
contributed to overall falling
deficit and debt figures, whereas
Growth (GDP in real terms, change in %)
in the case of Romania the
Bulgaria
2.3
5.4
4.1
4.8
4.3
primary balance rose in 2003 and
Romania
-1.2
2.1
5.7
5.0
4.9
was one of the elements why the
Turkey
-4.7
7.3
-7.5
7.9
5.8
debt and deficit levels did not fall
Inflation (CPI, change in %)
further.
Bulgaria
2.6
10.3
7.4
5.8
2.3
Romania
Turkey
3. ECONOMIC BACKGROUND
The overall improvement of
reported deficit and debt figures
in candidate countries occurred
against the background of
generally
favourable
and
3
45.8
64.9
45.7
54.9
34.5
54.4
22.5
45.0
15.3
25.3
6.7
19.8
-
5.2
14.2
-
Interest rate (5-year govt. bonds, % p. a.)
Bulgaria
Romania
Turkey
14.6
9.4
63.7
-
49.6
-
7.3
42.0
-
Source: Eurostat
However, strong growth of the economies and government revenues since then in all three countries
will lead most likely to better results than foreseen in March and than last year’s balances.
Consequently, the expected outcomes (see for example the European Commission Autumn 2004
forecasts) for 2004 will show significant improvements of government budgets in all three candidate
countries concerned vis-à-vis 2003 and also vis-à-vis the forecasts made in the notifications.
-4-
improving macroeconomic conditions in these countries.
Table 3 gives some key indicators of economic development in the three countries. In
particular, growth rates have continued to be relatively high in 2003, and interest rates fell
further. Inflation rates have come down considerably in Romania and in particular in
Turkey, thereby reducing the growth of nominal GDP, and have remained at relatively low
levels in Bulgaria.
4. QUALITY AND RELIABILITY OF THE NOTIFIED DEFICIT AND DEBT FIGURES
The March 2004 notifications have been assessed in detail in the country evaluations (see
below). The purpose of the country evaluations is to review the figures and to analyze the
reconciliation between the national budget figures and the notified figures. Also,
EUROSTAT provided for the country sections a summary assessment of the compliance of
the notified figures with the ESA 95 methodology. The country evaluations mention the
various areas where progress has been made and where further improvement is required.
From a methodological point of view, there has been progress in the statistical quality of
reported data in all countries. Yet, further efforts are needed in all three countries in order
to compile deficit and debt data fully compatible with ESA 95 standards. This has, among
others, to be ensured by a better co-operation of institutions within the respective countries,
by a compilation of financial accounts based on ESA 95 and improvements are needed in
explaining better the link between the deficit and change in debt, and in this way reducing
the statistical discrepancies.
Table 4 summarizes the specific findings per country in this respect.
Table 4: Main findings as regards quality of notified data
Bulgaria
Accruals recording
Explanations on the transition between net
borrowing/lending and change in debt
Classification of special funds and health sector institutions
Compilation of financial accounts
Romania
Explanations on the transition between net
borrowing/lending and change in debt
Quality and the timeliness of financial accounts
Turkey
Delimitation of general government sector
Accruals recording
Compilation of a full set of ESA 95 accounts for general
government sector and its sub-sectors
-5-
EVALUATION OF THE FISCAL NOTIFICATION OF BULGARIA
1. Key fiscal indicators reported
The main figures reported by the Bulgarian authorities to the European Commission in
March 2004 (compared to the figures reported in April 2003) are shown in Table 1 and
Charts 1 and 2.
Table 1 - Bulgaria: General government indicators and nominal GDP
% of GDP
Notification
Net lending (+) /
borrowing(-)
1999
2000
2001
2002
2003
2004
2003
-0.5
-0.5
0.2
0.2
-0.8
-0.6
-0.1
-0.7
-0.7
0.4
Primary net lending (+)
/ borrowing (-)
2004
2003
3.6
3.6
3.9
3.9
1.4
1.6
2.0
1.6
1.3
4.1
Gross debt
2004
2003
73.6
73.6
66.2
66.4
53.2
53.0
46.2
52.0
45.1
79.3
Gross fixed capital
formation
2004
2003
3.5
3.5
3.0
3.1
2.8
3.4
3.2
3.9
3.7
3.7
Nominal GDP growth
rate (%)
2004
2003
11.1
10.7
8.8
9.6
6.8
8.7
10.1
12.5
2004 (1)
(1) planned
According to this year’s notification, the general government’s net borrowing is expected to
remain below 1% of GDP in 2004. The main difference to last year’s figures is the outturn
for 2003, when the general government position actually ended almost in balance compared
to a deficit of 0.7% of GDP expected in April 2003 (Chart 1). In view of higher tax
revenues than expected and an increasing current account deficit, the government had
revised its cash budget target in the course of the year from a deficit of 0.7% of GDP to a
balance. Methodological changes have also influenced the underlying figures for 2001 and
2002, although the 2001 general
government balance remains the same.
Chart 1 Bulgaria:
Ge neral governme nt ne t
Interest payments have been reduced
borrowing (-) (% of GDP)
substantially from 2002 onwards and
0.5
Bulgaria has been recording a primary
2003 notification 2004 notification
surplus over the reporting period. This
is also the result of a decline in general
government gross debt from close to
0
80% of GDP in 1999 to 46% by the end
of 2003 (Chart 2). Following the sharp
depreciation of the US dollar against
the euro and the Bulgarian lev, the debt
-0.5
level in 2003 has fallen much more
than expected in the previous
notification, given the still considerable
amount of foreign government debt
denominated in USD. Gross fixed
-1
capital formation has moved in a
2000
2001
2002
2003
2004
corridor of between 3% and 4% of
-6-
GDP over the reporting period except
for 2003 when it was below 3% of
GDP, much lower than planned in
last year’s notification.
Chart 2
Ge ne ral gove rnme nt gross
de bt (% of GDP)
Bulgaria:
2003 notification
2004 notification
In
absolute
terms,
general
60
government gross debt was almost
unchanged in 2001 compared to
2000, decreased significantly in 2002
40
and 2003 and is expected to increase
in 2004 almost to its 2002 level. In
2003, the debt level was BGN 2400
million less than expected in last
20
year’s notification. The different
contributions to the change in debt,
expressed in percentage points of
0
GDP, from the primary surplus,
2000
2001
2002
2003
2004
interest and nominal GDP growth as
well as other contributions are
indicated in the upper part of Table 2. The primary budget surplus of 1.4% of GDP in 2002,
2.0% in 2003 and expected 1.3% in 2004 has a debt-reducing effect of the same size.
Nominal GDP growth and declining interest rates had strong debt-reducing effects in all
years, but less so in 2003 due to an unexpectedly low inflation rate. As can be seen in the
lower part of Table 2, other contributions are playing the most important role although with
different signs in 2003 and 2004.
These other contributions occur in three different ways. First, net acquisitions of financial
assets in the form of currency and deposits, securities, loans as well as shares and other
Table 2 - Bulgaria:
Contribution to changes in general government gross debt and gross debt ratio
Change in gross debt ratio
to which contribution of …
• Primary balance
• Interest and nominal GDP growth
• Other
2002
GDP % pts*
-12.9
2003
GDP % pts*
-7.0
-1.4
-3.1
-8.3
2004 (1)
GDP % pts*
-1.1
-2.0
-1.3
-3.8
-1.3
-2.1
2.4
mio. BGN % of GDP mio. BGN % of GDP mio. BGN % of GDP
General government net borrowing
+ Other contributions (2)
= Change in general government gross debt
247
-2699
-2452
0.8
-8.3
-7.6
42
-1310
-1268
0.1
-3.8
-3.7
284
895
1179
0.7
2.4
3.1
General government gross debt
17201
53.2
15932
46.2
17112
45.1
* differences of ratios to GDP in year t to ratios to GDP in year t-1
(1) planned
( 2) Net acquisition of nominal assets, appreciation/depreciation of foreign currency debt and remaining statistical adjustments
equity play a minor role. For 2004, a stronger use of loans is envisaged, whereas previous
years saw a higher influence of debt management operations and privatisation. Second,
adjustments following changes in the value of foreign-currency debt are by far the most
important contribution. These changes result from swap activities and, most importantly
because of the high importance of USD-denominated debt, from exchange rate movements.
-7-
In 2002 and 2003, the depreciation of the USD against the EUR and the BGN was 3.6% and
20.8% on average. These factors contributed to a decrease in nominal debt of BGN 1805
million in 2002 and of BGN 1764 million in 2003 whereas the notification expects a
reversal for 2004 with an increase of BGN 683 million. Third, statistical discrepancies are
also considerable, although diminishing in size over the reporting period.
2. The macroeconomic context
The prudent fiscal stance taken in the reporting period has both contributed to and benefited
from a favourable macroeconomic environment. Domestic demand was very strong, based
on rising net income, high fixed investment and very high bank credit growth. However,
given the slow economic activity in the EU as its main trading partner, this triggered a
substantial deterioration of the external balance. As indicated in Table 3, real GDP growth
has remained high at 4.3% in 2003. Average consumer price inflation declined from 5.8%
in 2002 to 2.3% in 2003 but, compared to one year earlier, has been moving upwards to
5.6% at the end of 2003 as a result of increases in energy and food prices. The current
account deficit in 2003 was at 8.6% of GDP, with a trade deficit of 12.5% of GDP, although
figures include net errors and omissions of 2% of GDP. Net foreign direct investment was
also exceptionally high at 7% of GDP and the central bank’s foreign currency reserves
continued to increase. The unemployment rate in 2003 decreased by 4½ percentage points
as a result of a statistical effect of a stronger enforcement of labour contracts, newly
introduced active labour market policies and net job creation in the private sector. Total
foreign debt decreased to below 60% of GDP at the end of 2003.
Table 3 - Bulgaria: Main economic trends
annual averages
Growth
Inflation
Unemployment
Current account
Interest rate
Exchange rate
1999
2000
2001
2002
2003
2.3
2.6
5.4
10.3
16.9
-5.6
9.4
1.953
4.1
7.4
20.3
-7.3
7.3
1.948
4.8
5.8
18.2
-5.6
6.7
1.949
4.3
2.3
13.7
-8.6
5.2
1.949
GDP in real terms, change in %
CPI, change in %
LFS, % of labour force
balance, % of GDP
5-year govt. bonds, % p. a.
BGN/EUR
-4.8
14.6
1.956
Source: Eurostat
Tax revenues in 2003 were much higher than expected, mostly because of high corporate
profits and customs revenues, and a cash surplus of 2.8% of GDP accumulated until
October which was subsequently spent until the end of the year. The three budget buffers,
which give fiscal policy considerable flexibility and result in a spending profile biased
towards the end of the year, allow a better targeting of the end-year deficit: First, a
contingency reserve for natural disasters and structural reforms; second, an “88%-rule” to
limit discretionary spending in the first three quarters so that the rest is only spent if the
budget is running according to plan; third, the fiscal reserve account of 11.2% of GDP at the
end of 2003, most of which is held at the central bank and only to be used in exceptional
circumstances. The government continued its active debt management strategy with the
objective of reducing risks by gradually shifting from a denomination in US dollar towards
a denomination in euro, from short-term to long-term maturity, from floating into fixed
interest rate bonds and from foreign to domestic financing.
-8-
3. Methodological issues
Main challenges
The data have been calculated by applying as far as possible accounting and definitional
requirements (ESA 95) used for the fiscal surveillance of EU Member States. Some
methodological progress has been made in this year’s notification but further significant
improvements are required in particular in the use of accrual-based data as the 2004
notification still relies predominantly on cash-based figures. It is expected that the next
notification will include accrual-based figures, in particular for taxes and interest payments.
This should make the government figures more stable and more comparable between
successive years.
Transposition of national budget balances into ESA 95
Table 4 quantifies the transposition of the national budget balance into the ESA 95 general
government net borrowing/net lending definition. The first line shows the actual (and for
2004 planned) figures of the most prominent budget balance which is in Bulgaria the State
budget as approved by the Parliament (including the central budget, the ministries’ budgets,
the budget of the National Audit Office and the judiciary authorities’ budget).
Table 4 - Bulgaria:
Transposition of national budget balances into ESA 95 net lending (+)/borrowing (-)
2002
2003
2004 (1)
mio.
BGN
% of
GDP
mio.
BGN
% of
GDP
mio.
BGN
% of
GDP
6
0.0
-101
-0.3
-237
-0.6
+ adjustment to central government net lending
-309
-1.0
119
0.3
94
0.2
= Central government net lending (S.1311)
-303
-0.9
18
0.1
-143
-0.4
+ Local government net lending (S.1313)
+ Social security net lending (S.1314)
-31
88
-0.1
0.3
-66
7
-0.2
0.0
-78
-63
-0.2
-0.2
= General government net lending (S.13)
-247
-0.8
-42
-0.1
-284
-0.7
Most prominent national budget balance (2)
(1) planned
(2) State budget balance (the State budget includes the central budget, the ministries’ budgets, the budget of the
National Audit Office and the judicial authorities’ budget)
Line 2 of Table 4 adjusts for four differences between the State budget balance and the
ESA 95 concept of central government net borrowing. The first adjustment applies to net
loans granted to non-financial public enterprises and payments for their defaulted debts
which were accounted as expenditure in the consolidated budget until 2001. In the
subsequent budgets an adjustment for these transactions was no longer necessary. Second,
until 2002 an adjustment is needed for central bank profits which are recorded on a cash
basis as budget revenues one year after they accrued; they are recorded on an accrual basis
from 2003 onwards. Third, an adjustment is needed for the net lending/borrowing position
of bodies which are not part of the central government but should be accounted as such
under ESA 95 methodology. The difference between the State budget balance and central
government net borrowing is mostly explained by this adjustment for the balance of the
extra-budgetary accounts and funds, which fluctuated between a net lending of BGN 202
million in 2001 to a net borrowing of about BGN 294 million in 2002; net lending in 2004
is expected to be smaller than in 2003. The State enterprise for managing environmental
protection activities, which came into operation at the beginning of 2003, had a net
-9-
borrowing requirement of BGN 43 million in 2003. Fourth, receipts of BGN 236 million in
2001 from the sale of a GSM license are taken into account. Regarding the other sectors of
general government, local governments ran a small surplus in 2001 and small but increasing
deficits since then. The balance for the Social Security sub-sector shows a surplus with a
declining trend turning into an expected deficit in 2004.
Stability of data4
Compared with the notification submitted in April 2003, the data for 2000 are unchanged.
For 2001 the local government surplus has been lowered by around BGN 60 million (0.2%
of GDP), while central government deficit has been reduced by a similar amount. This is
explained by a correction in the classification of certain extra-budgetary accounts from
central to local government. For 2002, the net borrowing of general government is now 0.76% of GDP compared with -0.65% reported in April 2003, an upward revision to net
borrowing of extra-budgetary funds being largely offset by the recording of an adjustment
for transfer of central bank profits. General government net borrowing for 2003 is now 0.12% of GDP compared with a forecast a year ago of -0.75%. Local government and
social security fund balances are worse than forecast, but central government ended in a
small surplus (0.05% of GDP) compared with a forecast net borrowing of 0.9%. This
difference is reflected in the reported public balance for central government. The lower
level of GDP in 2003 compared with the forecast a year ago should be seen in the context of
a downward revision to GDP for 2002.
Total consolidated debt for 2003 is much lower than previously expected (46.2% of GDP,
compared with a forecast 52% in the previous notification), mainly because of the change in
value of foreign debt caused by the US dollar’s depreciation against the Lev. As a result the
outstanding value of long-term securities was revised downwards.
The number of extra-budgetary funds and accounts has been considerably reduced from 28
in 1999 to 7 in 2003, accounting for sharp variations in their total net lending /borrowing.
Institutional changes in the healthcare system are believed to explain variations in the
balance on social security funds over time. Eurostat is not currently in a position to verify
whether the accounting treatment of special funds and health sector institutions is made in
full compliance with ESA 95. Local government debt is still not provided for 2000 and
2001, but debt of social security funds for these years are now included in the notification.
Deficit and debt methodology
In the calculation of net borrowing/lending there are no adjustments to be made for
financial transactions for 2002 and 2003, as none are included in the working balances after
2001. The notification is based on cash data, except for central bank profits calculated on
the basis of time-adjusted cash. The Ministry of Finance is finalising accrual-based tax data
and interest expenditure on accrual basis from the year 2000 onwards, but these are not yet
included in the notification. The accounting treatment of uncollectible taxes is being
examined by Eurostat and a change in accounting may be required for future notifications.
The adjustment shown for transfer of central bank profits to government has been made,
according to the statistical authorities, on the basis of time-adjusted cash. However, the
ESA 95 methodology states that dividends should be recorded when they are due to be paid.
The figures are therefore likely to be adjusted for a future notification.
Net
borrowing/lending of the Bank Consolidation Company has been provided for 2001 and
2002. An estimate is expected to be made for 2003.
4
The remainder of this section has been provided by Eurostat.
- 10 -
In the tables showing the transition from net borrowing/lending to the change in debt for
each sub-sector of government, there are numerical differences which are recorded under
'statistical discrepancies' (in addition to those discrepancies already shown). There are
significant statistical discrepancies in the reconciliation of general government net
borrowing /lending with the debt for all years, especially for 2002 when the difference is
equivalent to 1.6% of GDP. According to the authorities part of the discrepancy in 2002 is
because of 'non-cash' transactions, involving exchanges of Brady bonds with other bonds.
These transactions have not been fully shown under the appropriate headings in table 3.
Exchange operations involving Brady bond collateral largely account for the sharp
variations in net acquisition of securities over time. Part of the statistical discrepancy in
2003 relates to the reclassification of government guaranteed loans into the debt of general
government. Once detailed information is received from the Bulgarian authorities
regarding the precise nature of government guarantees granted on loans and actual calls
made on those guarantees, Eurostat will recommend a classification of these loans. The
national authorities claim that there are no outstanding government trade credits and
advances, nor any outstanding amount related to financing of public undertakings.
Little is known about the specific re-classifications of health sector institutions into, or out
of, the general government sector. Eurostat is therefore currently not in a position to verify
whether the accounting treatment of special funds and health sector institutions is made in
full compliance with ESA 95.
Gross Domestic Product
Concerning the quality of GDP data, the conceptual and practical compliance with ESA 95
has continuously improved, but special attention needs to be paid to the exhaustiveness of
national accounts data and related data sources before full compliance can be certified.
4. Conclusions
The figures reported provide evidence of Bulgaria’s continuing commitment to high fiscal
discipline by keeping the general government deficit below 1% of GDP and nearly balanced
in some years. The primary surplus, sustained economic growth and other factors have
contributed to the substantial decline in the debt-to-GDP ratio from 74% of GDP in 2000 to
expected 45% of GDP in 2004. The prudent fiscal stance taken in the reporting period has
both contributed to and benefited from a favourable macroeconomic environment.
However, given the very strong domestic demand and the slow economic activity in the EU
as its main trading partner, this could not prevent the substantial deterioration of the current
account in 2003.
From a methodological point of view, the Bulgarian authorities are making some progress
in improving the statistical quality of the deficit and debt notification. Greater efforts are
required in particular on the consistent production of accrual-based data, on giving plausible
explanations on the transition between net borrowing / net lending and the change in debt,
and on the classification of special funds and health sector institutions. Compilation of
financial accounts, which is still at an early stage, is expected to enhance the robustness of
the notified data.
- 11 -
EVALUATION OF THE FISCAL NOTIFICATION OF ROMANIA
1. Key fiscal indicators reported
The main figures reported by the Romanian authorities to the European Commission in
March 2004 (compared to the figures reported in April 2003) are shown in Table 1 and
Charts 1 and 2.
Table 1 - Romania: General government indicators and nominal GDP
% of GDP
Notification
Net lending (+) /
borrowing(-)
1999
2000
2001
2002
2003
2004
2003
-4.4
-4.6
-3.5
-3.3
-2.0
-2.2
-2.0
-2.5
-3.0
-4.5
Primary net lending (+)
/ borrowing (-)
2004
2003
-0.7
-0.3
-0.3
0.0
0.2
0.2
-0.4
-0.4
-1.4
0.9
Gross debt
2004
2003
23.2
23.1
23.3
22.7
21.8
23.7
23.5
24.0
23.9
23.9
Gross fixed capital
formation
2004
2003
1.9
1.9
2.4
2.4
3.1
2.5
3.2
2.6
3.0
2.1
Nominal GDP growth
rate (%)
2004
2003
45.3
45.2
29.5
29.6
25.0
23.6
19.2
47.3
2004 (1)
According to this year’s notification, after a deficit of 4.4% of GDP in 2000 (see Graph 1),
general government net borrowing declined and was more than halved by 2002 when the
budgetary target for the year was overachieved. The main difference to last year’s figures is
the outturn for 2003 as a slight undershooting of the deficit target again occurred due to
strong growth in tax revenue and further savings on interest payments. Largely as a result
of this better than expected outturn, budgetary plans imply an increase in the general
government deficit in 2004. In line with Romania’s Pre-Accession Economic Programme
the deficit is expected to reach 3.0% of GDP on the back of more subdued revenue growth
and a further solid expenditure growth,
part of which is assigned to maintaining
fixed capital formation at a level around
Chart 1 Romania:
Gene ral gove rnment ne t
3% of GDP after an upwards revision of
borrowing (-) (% of GDP)
1
reported figures for 2002 and 2003.
2003 notification
Interest payments have been reduced
substantially over recent years, mainly
due to a decline in market refinancing
rates and improved debt management.
This
witnesses
a
significant
deterioration of the primary balance
which after being in surplus in 2002 for
the first time since 1999 returned to a
deficit position in 2003. Given the
planned
stability
of
interest
expenditures as a percentage of GDP
this year and the larger general
government deficit, the primary balance
is expected to worsen by 1% of GDP in
2004 notification
0
-1
-2
-3
-4
-5
2000
- 12 -
2001
2002
2003
2004
2004.
In absolute terms, general government gross debt has increased steadily since 2000, and will
continue to do so in 2004. In 2003, the debt level was, however, ROL 30000 billion, or
approximately 7%, lower than expected in last year’s notification. Accordingly, the debtto-GDP ratio has declined noticeably by
1.5% of GDP in 2003, reaching 21.8%
Chart 2 Romania: Ge ne ral gove rnme nt gross
of GDP (Chart 2). Strong nominal GDP
de bt (% of GDP)
growth and falling interest payments
30
more than counterbalanced the impact
of other factors, which tended to
increase the debt ratio like the
depreciation of the currency. Moreover,
20
given the still considerable amount of
foreign government debt denominated
in USD, the broadly stable Romanian lei
against the US dollar over 2003 has
contributed to the decline in the debt
10
ratio. As a result of slowing nominal
GDP growth and the envisaged
deterioration in the primary balance,
2003 notification 2004 notification
however, the declining trend in
0
Romania’s
moderate
general
2000
2001
2002
2003
2004
government debt-to-GDP ratio is
projected to come to an end. For 2004,
an increase by 1.7% of GDP is assumed.
The different contributions to the change in debt, expressed in percentage points of GDP,
are indicated in the upper part of Table 2. The primary budget surplus of 0.2% of GDP in
Table 2 - Romania:
Contribution to changes in general government gross debt and gross debt ratio
Change in gross debt ratio
to which contribution of …
• Primary balance
• Interest and nominal GDP growth
• Other
2002
GDP % pts*
+0.1
2003
GDP % pts*
-1.5
-0.2
-3.1
3.4
2004 (1)
GDP % pts*
+1.7
0.4
-3.1
1.2
1.4
-1.9
2.2
mio. ROL % of GDP mio. ROL % of GDP mio. ROL % of GDP
General government net borrowing
+ Other contributions (2)
= Change in general government gross debt
30,353
50,844
81,196
2.0
3.4
5.4
37,216
22,484
59,699
2.0
1.2
3.2
67,453
50,378
117,831
3.0
2.2
5.2
General government gross debt
352,435
23.3
412,135
21.8
529,966
23.5
* differences of ratios to GDP in year t to ratios to GDP in year t-1
(1) planned
(2) Net acquisition of nominal assets, appreciation/depreciation of foreign currency debt and statistical adjustments
2002 had a debt-reducing effect of the same size, while the primary deficit of 0.4% of GDP
in 2003 and expected 1.4% of GDP in 2004 are a main factor in explaining the changes in
gross debt. Nominal GDP growth and declining interest payments had strong debt-reducing
effects in all years, but less so in 2004 due to the expected decline in inflation and higher
interest payments in absolute terms.
- 13 -
As can be seen in Table 2, other contributions play an important role in all years. These
occur in three different ways. First, net acquisitions of financial assets in the form of
currency and deposits, securities, loans as well as shares and other equity play a large role.
This includes privatisation receipts and credits to enterprises. Second, adjustments
following changes in the value of foreign-currency debt give an important contribution,
resulting mainly from the depreciation of the national currency against other currencies,
which in 2003 contributed to an increase in nominal debt of 1.1% of GDP. Third, statistical
discrepancies are considerable, although having diminishing markedly.
2. The macroeconomic context
The relatively prudent fiscal stance taken in the reporting period has both contributed to and
benefited from a favourable macroeconomic environment. Domestic demand was very
strong, based on higher disposable incomes, sustained growth in investments and high
credit growth to the non-government sector. In 2003, this triggered strong import growth,
and with export performance at the same time being less vigorous compared to previous
years, net exports turned strongly negative. As indicated in Table 3, real GDP growth
remained high at 4.9% in 2003. Average consumer price inflation remained on its
downward path, declining from 22.5% in 2002 to 15.3% in 2003, but remained relatively
high. As a result of the worsened trade balance, the current account deficit widened
significantly over 2003, closing at 5.8% of GDP. Net foreign direct investment picked up
moderately to 2.7% of GDP. The unemployment rate declined in 2003 despite lay-offs in
state-owned enterprises. Job creation seems to be strong in the private sector, but total
employment decreased, which could be a sign of retreating labour market participation.
Total foreign debt remained stable at 34% of GDP at the end of 2003.
Table 3 - Romania: Main economic trends
Growth
Inflation
Unemployment
Current account
Interest rate
Exchange rate
annual averages
1999
2000
2001
2002
2003
GDP in real terms, change in %
-1.2
45.8
6.3
-4.3
63.7
2.1
45.7
6.8
-3.4
49.6
5.7
34.5
6.6
-5.0
42.0
5.0
22.5
7.4
-3.6
19.8
4.9
15.3
6.6
-5.8
14.2
16310.3
19934.6
26010.3
31248.3
37547.3
CPI, change in %
LFS, % of labour force
balance, % of GDP
5-year govt. bonds, % p. a.
ROL/EUR
Source: Eurostat, National Bank of Romania
The very prudent fiscal performance over the first eleven months of 2003 with a cash
balance of only -0.8% of GDP was reversed and turned pro-cyclical towards the end of the
year as revenue gains were largely spent. Strong growth in tax revenue and further savings
on interest payments resulted, however, in a slight undershooting of the deficit target for
2003. The explanation for this turnaround seems to be found in an insufficient expenditure
control against the background of a better than expected development in revenues. The
progress on the revenue side was both due to cyclical factors, resulting in very high growth
rates for excises, customs duties, VAT and profit tax, but also improved collection such as
for the income tax and the social security contributions. The government continued its
active debt management strategy with the objective of gradually extending the average
maturity of debt, improve risk management by the use of both fixed and variable interest,
and increasingly develop domestic financing, most notably by establishing pricing
benchmarks over a three-year maturity horizon for domestic paper. However, external
financing in 2003 remained above the level envisaged, and the maturity of government
bonds sold domestically was shorter than intended.
- 14 -
3. Methodological issues
Main challenges
The data have been calculated by applying as far as possible the accounting and definitional
requirements (ESA 95) used for the fiscal surveillance of EU Member States. Progress has
been made in this year’s notification in moving from a cash basis to an accrual basis.
Moreover, statistical discrepancies have diminished, albeit being still significant. However,
further progress is required, such as an improved accounting of EU transfers, improved
explanation of the link between the deficit and change in debt and clearer classification of
units, in order to fully meet the ESA 95 requirements.
Transposition of national budget balances into ESA 95
Table 4 quantifies the transposition of the national budget balance into the ESA 95 general
government net borrowing/net lending definition. The first line shows the actual (and for
2004 planned) figures of the most prominent budget balance which is in Romania the cash
State budget as approved by the Parliament.
Table 4 - Romania:
Transposition of national budget balances into ESA 95 net lending (+)/borrowing (-)
2002
mio. ROL
Most prominent national budget balance (2)
+ adjustment to central government net lending
= Central government net lending (S.1311)
2003
% of
GDP
-75,119.00 -4.97
39,613.70
2.62
mio. ROL
2004 (1)
% of
GDP
-66,105.20 -3.50
30,412.20
1.61
mio. ROL
% of
GDP
-100,591.60 -4.46
28,630.30
1.27
-35,505.30 -2.35
-35,693.00 -1.89
-71,961.30 -3.19
+ Local government net lending (S.1313)
+ Social security net lending (S.1314)
-433.90 -0.03
5,586.60 0.37
-0.90 0.00
-1,521.80 -0.08
-316.10 -0.01
4,824.10 0.21
= General government net lending (S.13)
-30,352.60 -2.01
-37,215.70 -1.97
-67,453.40 -2.99
(1) planned
(2) Balance of the State budget and special funds, including balance of EU
Line 2 of table 4 adjusts for some major differences between the State budget balance and
the ESA 95 concept of central government net borrowing. The first adjustment applies to
net loans granted to non-financial public enterprises, repayment of their defaulted debts and
receipts from privatisation. Financial transactions, like lending operations or sale of public
equity, need to be excluded while, conversely, any non-financial transaction excluded in the
calculation of the national budget balance needs to be included. With privatisation revenues
usually financing early debt redemption of a comparable magnitude, the exclusion of
government lending activities and debt amortisation account for most of the adjustment.
Each year during the reporting period, this implied a positive adjustment, which peaked at
2.2 % of GDP in 2002. Second, since the ESA 95 balance is an accrual concept, another
adjustment is needed since the most prominent national budget balance is recorded on a
cash basis. The resulting correction tends to further reduce the deficit. For the first time,
taxes and social contributions are apparently fully recorded on an accrual basis in
compliance with ESA 95. Interest payments are only recorded on an accrual basis from
2001 onward. Third, a smaller adjustment is needed for the net lending/borrowing position
of bodies which are not part of the central government but should be accounted as such
under ESA 95 methodology, such as public institutions partially or fully financed from
extra-budgetary incomes. Finally, an adjustment is made relating to capital transfers by the
- 15 -
government to the banking sector following the purchase of non-performing loans by the
defeasance body (AVAB). The adjustment was important early in the reporting period, but
has since 2002 been negligible. Regarding the other sectors of general government, local
governments ran broadly balanced budget over the reporting period. The balance for the
Social Security sub-sector has remained in surplus in all years except 2003, where increased
pension allocations and compensation packages for laid off workers stretched the finances.
Stability of data5
The data for 2000-2001 are reported as final but should be considered half-finalised.
Compared to the previous notification of March 2003 revisions have been made for net
borrowing/net lending of general government for 2000-2003. Improvement of the deficit
occurred for the years 2000, 2002 and 2003, while a worsening of the deficit was recorded
for 2001. The revisions are mainly due to taking into account calculation of interest on
accrued basis for the domestic debt. Until now interest payments had been recorded on
accrued basis only for external debt.
Concerning 2003 the improvement of the deficit of general government compared to the
previous notification is due to a better financial outturn for central government (-2.9% of
GDP in the March 2003 notification compared to –1.9% now). The authorities explain this
improvement in terms of both increased revenue and lower expenditure. However for the
local government and social security sub-sectors, a deficit was recorded in 2003 compared
to a surplus forecast a year ago. The level of general government debt for 2003 has been
revised downwards (21.8% of GDP against 23.7% of GDP). This reduction is reflected in a
sharp decrease in the level of short-term loans partly offset by an increase in the long-term
loans.
Concerning gross fixed capital formation, upward revisions for 2002 and 2003 are recorded
due, it is said, to the availability of more detailed data sources. The downward trend in
interest payments for 2000-2003 is due to lower interest rates on the domestic market and to
borrowings in foreign currency at a lower cost. Compared to the previous notification only
slight revisions concerning GDP occurred.
Deficit and debt methodology
Delimitation of the general government sector seems to be generally in line with ESA 95
methodology. Nevertheless an analysis in depth of the units to be included or excluded
from the general government sector should be undertaken in the near future. It is not clear
whether the National Road Agency, which is treated as being outside government sector,
should not be reclassified within it.
Moreover the Statistical Office is in the process of revising the classification of the selffinancing units subordinated to ministries and other agencies of central and local
government. The final results are expected during the summer and could lead to the
reclassification of some units. The Eurostat decision on private public partnerships (PPP) is
being analysed by the Statistical Office and its impact on general government accounts will
be known by the end of the year.
The working balance of central government includes the outturn of the State Budget as well
as the special funds, and is compiled on a cash basis. The State Treasury fund and public
institutions financed from extra-budgetary incomes are not included, but an adjustment is
recorded under ‘net borrowing/net lending of other central government bodies’. The
5
The remainder of this section has been provided by Eurostat.
- 16 -
working balance includes EU grants received and paid. This cash balance does not have any
impact on the deficit as the amounts received are equal to the amounts paid.
Financial transactions included in the public accounts appear to be treated correctly in table
2. Privatisation receipts are separately identified. However a significant adjustment is made
concerning the item 'loans granted' while very little is recorded under the item 'loans
repayments'. According to the authorities the amounts recorded under loans granted include
repayments of loans by government. It should be investigated if these repayments concern
loans issued by the government or loans received by companies with a state guarantee and
paid back by government.
The transactions concerning taxes and social contributions are recorded using the “time
adjusted cash” method and seem to comply with ESA 95. However, more investigation
should be made into the issue of taxes and social contributions assessed but never collected.
Moreover, the method for recording of taxes used for data reported under the ESA 95
transmission programme is apparently on an assessment basis, and results in different
figures for government net borrowing.
The item 'other adjustments' relates to capital transfers made by government to the banking
sector following the purchase of non-performing loans by the defeasance body (AVAB).
The method of recording seems to comply with ESA 95. However the issue of capital
transfers in general should be examined further.
Reconciliation between general government net borrowing/net lending and the change in
gross debt is rather difficult to assess due to lack of stability in the data and to the fact that
financial accounts have been compiled up to 2000 only. No government debt is held by
other general government sub-sectors.
This means that the general government
consolidated debt is equal to the general government non-consolidated debt. It is worth
noticing that the government systematically issues and redeems debt below par. The main
factor explaining the change in government debt for 2000-2003 is appreciation of foreign
currency debt. However important revisions occurred compared to the previous notification
due to the fact that publicly guaranteed debt is now taken into account. This matter needs
further investigation. The statistical discrepancies related to local government and social
security funds are quite important in terms of the change and level of debt of these subsectors.
There have been no UMTS licence allocations in Romania. The government has not been
involved in any swap or FRA arrangements.
Gross Domestic Product
Concerning the quality of GDP data, the conceptual and practical compliance has
continuously improved but further efforts are needed as concerns data sources and treatment
of financial services there is still some distance to go before full compliance with the 'acquis
communautaire' can be certified. Exhaustiveness problems still exist, particularly in
relation to a not fully reliable business register and the self-employed and the robustness of
the adjustments.
4. Conclusions
The figures reported provide evidence of Romania’s commitment to stronger fiscal
discipline by keeping the general government deficit within prudent limits. The prudent
fiscal stance in the reporting period both contributed to and benefited from a favourable
- 17 -
macroeconomic environment. The primary deficit, however, is increasing rather rapidly
and contributes considerably to the expected increase in the debt-to-GDP ratio in 2004.
From a methodological point of view, the authorities have made some progress towards
satisfying EU reporting requirements for government deficit and debt, such as recording
taxes and social contributions and interest payments on an accrual basis. However, further
institutional co-operation is necessary in order to improve the quality and the reliability of
the data. In particular, improvement is needed in explaining better the link between the
deficit and change in debt, and in this way reducing the statistical discrepancies. Efforts to
improve the quality and the timeliness of financial accounts are important in this respect.
- 18 -
EVALUATION OF THE FISCAL NOTIFICATION OF TURKEY
1. Key fiscal indicators reported
The main figures reported by the Turkish authorities to the European Commission in March
2004 (compared to the figures reported in April 2003) are shown in Table 1 and Charts 1
and 2.
Table 1 - Turkey: General government indicators and nominal GDP
% of GDP
Notification
Net lending (+) /
borrowing(-)
1999
2000
2001
2002
2003
2004
2003
-6.1
-5.8
-29.8
-26.9
-12.7
-10.0
-8.7
-11.3
-8.0
-18.4
Primary net lending (+)
/ borrowing (-)
2004
2003
7.9
8.2
-2.7
0.3
7.1
9.8
9.6
7.2
7.6
3.4
Gross debt
2004
2003
105.2
105.4
94.9
95.0
87.1
86.3
85.0
67.4
57.4
57.6
Gross fixed capital
formation
2004
2003
5.3
5.2
5.1
4.6
3.9
4.4
3.9
5.0
5.1
5.2
Nominal GDP growth
rate (%)
2004
2003
43.2
43.2
54.7
54.7
30.8
28.5
17.1
60.9
2004 (1)
(1) planned
The 2004 notification illustrates the continued process of fiscal consolidation after major
economic shocks (the earth quake in 1999 and the financial crisis in 2001) had resulted in a
dramatic deterioration of public sector finances. In particular, the 2001 financial crisis had
a major impact on Turkey’s public finances: the general government deficit rose from 6.1%
of GDP in 2000 to 29.8% in 2001 and the debt ratio increased from 57.4% of GDP in 2000
to 105.2%. In order to improve the sustainability of public finances, the Turkish authorities
have been targeting substantial primary surpluses, which were above 7% of GDP during
most of the reporting period. Only in
Chart 1 Turkey:
Ge neral governme nt ne t
2001, substantial expenditures to cover
borrowing (-) (% of GDP)
the losses of state banks led to a primary
10
deficit of 2.7% of GDP. Another
2003 notification 2004 notification
important feature of Turkey’s public
finances is the high burden of interest
0
payment. Real domestic interest rates
are high, reflecting high economic
-10
uncertainty and a tight domestic capital
market. Furthermore, interest rates are
closely related to exchange rate
-20
fluctuations and changes in market
sentiment. As a result, financing costs
are highly pro-cyclical and are difficult
-30
to predict. During the reporting period,
the financing costs increases sharply
-40
from 14% of GDP in 2000 to 27% in
2000
2001
2002
2003
2004
2001. Since then, the interest rate
burden has declined to 18.3% of GDP in
- 19 -
2003. For 2004, interest payments amounting to 15.6% of GDP are foreseen. Gross fixed
capital formation is relatively low and has further declined from slightly above 5% of GDP
during 2000-2002, to 3.9% in 2003. For 2004, a similar level is expected. Due to
decreasing inflation, nominal GDP growth decelerated from 54.7% in 2002 to 30.8% in
2003. For 2004, a further slowdown in nominal growth is assumed. However, the declining
nominal growth mainly reflects declining inflationary pressures, while real output growth
remains strong.
Compared to the 2003 notification, the main change is a better alignment of the treatment of
accrued interest with ESA 95 standards. This measure led to a deterioration of the primary
balance by nearly 3 percentage points in 2001 and 2002. For 2003, more detailed
information is now included in the notification, resulting in a primary surplus ratio which is
about 2½ percentage points higher than the estimate notified last year.
After the sharp increase in the debt
level from 57.4% in 2000 to 105.2% in
2001, the debt-to-GDP ratio has been
declining, reaching 94.9% in 2002 and
87.1% in 2003. For 2004, a further
reduction in the debt ratio by 2
percentage points is planned (Chart 2).
Chart 2
Turke y:
Ge ne ral gove rnme nt gross
de bt (% of GDP)
100
80
Table 2 takes a closer look at the
60
factors behind the debts dynamics. The
upper part of the table analyses the
40
annual change in the debt-to-GDP ratio
and decomposes these changes into
several underlying factors: the impact
20
of the primary surplus, the combined
2003 notification 2004 notification
effect of interest rates and nominal
0
GDP growth and all other factors.
2000
2001
2002
2003
2004
During 2002-2004, the debt-to-GDP
ratio declined substantially, by 10.4
percentage points in 2002, by 7.8% percentage points in 2003 and for 2004 a further
reduction by 2.1 percentage points is expected for 2004. The most noteworthy aspect in this
context is the important, but declining role of nominal GDP growth and interest payments
for the dynamics of debt ratio. As can be seen in Table 2, “interest and GDP growth”
reduced the debt ratio by 17.5 and 4.1 percentage points in 2002 and 2003. In 2004, this
category is expected to have a debt-ratio increasing impact of 2.9 percentage points. The
main underlying reason for this development is the sharp decline in nominal GDP growth
rates due to decreasing inflationary pressures. As a result, the debt reducing impact of
nominal GDP growth declined from 37.2 percentage points in 2002 to 22.4 percentage
points in 2003. For 2004, a debt-ratio reducing effect of 12.7 percent of GDP is expected.
In contrast to the debt-ratio reducing contribution of nominal GDP growth, the debt-ratio
increasing impact of interest payments is declining at a slower pace. In 2002 and 2003, the
debt service added 19.8 and 18.3 percentage points to the debt ratio. In 2004, the
contribution of debt payments is expected to continue to decline to 15.6% of GDP. As a
result, the net effect of nominal GDP growth and interest rate will change from a debt
reducing factor in 2002 and 2003 to a debt increasing factor in 2004. As a result, the
primary surplus is becoming an increasingly important factor for the reduction in the debt
ratio: In 2002 and 2003, the primary surplus reduced the debt-to-GDP ratio by 7.1 and 9.6
percentage points, while in 2004, it is expected to reduce the debt ratio by 7.6 percentage
points.
- 20 -
Table 2 - Turkey:
Contribution to changes in general government gross debt and gross debt ratio
Change in gross debt ratio
to which contribution of …
• Primary balance
• Interest and nominal GDP growth
• Other
2002
GDP % pts*
-10.4
2003
GDP % pts*
-7.8
2004 (1)
GDP % pts*
-2.1
-7.1
-17.5
14.1
-9.6
-4.1
5.9
-7.6
2.9
2.5
trillion TRL % GDP
trillion TRL % GDP
trillion TRL % GDP
General government net borrowing
+ Other contributions (2)
= Change in general government gross debt
35,055
38,994
74,049
12.7
14.1
26.8
31,509
21,221
52,730
8.7
5.9
14.6
34,030
10,765
44,795
8.0
2.5
10.6
General government gross debt
261,824
94.9
314,554
87.1
359,349
85.0
* differences of ratios to GDP in year t to ratios to GDP in year t-1
(1) planned
(2) Net acquisition of nominal assets, appreciation/depreciation of foreign currency debt and remaining statistical adjustments
The lower part of table 2 presents the absolute changes in the debt level and differentiates
between the impact of general government net borrowing and other contributions. This
latter category contains 3 different groups of variables which have an impact on the debt
level, besides the general government borrowing. The first group contains adjustments for
net acquisitions of financial assets in the form of currency and deposits, securities, loans as
well as shares and other equity. The most significant item in this group were the loans to
the Savings Deposit Insurance Fund, increasing the debt ratio by 3.1 percentage points in
2000 and by 7.5 percentage points in 2001. In 2003, additional loans to the SDIF to cover
the losses of the Imar bank increased the debt ratio by 2% of GDP. The category “other
financial assets” raises the debt by another 3.4-4.5% of GDP. Given the significance of
those transactions, a more detailed explanation would have been helpful. The second group
summarises changes in the value of existing assets and loans, or foreign exchange assets
and liabilities. In this group, the most significant impact came from exchange rate
fluctuations, increasing the debt by 18.2 percentage points in 2001 and by 6.2 percentage
points in 2002. In 2003, the appreciation of the currency decreased the debt by 3.2
percentage points. Another important factor were issuances below par, increasing the debt
by 4.9, 4.5 and 5.6% of GDP during 2000-2002. In 2003, this debt increasing item
amounted to 7.9%. Besides the general government net borrowing, this constituted the
most important single factor for the change in the general government debt. Finally,
statistical discrepancies increased the debt by at least 2½% of GDP during the reporting
period. In 2001, statistic discrepancies amounted to even 9.7% of GDP. In 2002, the
general government deficit counted for 12.7% of GDP, while other adjustments accounted
for 14.1% of GDP. In 2003, the impact of other adjustments on the debt level declined,
amounting to only 5.9% of GDP, while the general government deficit was 8.7% of GDP.
In 2004, the Turkish authorities expect a deficit of 8% of GDP, while other adjustments
affecting the debt level are expected to amount to 2.5% of GDP only. Overall, the notified
data indicates a continuous decline of debt relevant factors, which are not related to the
general government borrowing.
2. The macroeconomic context
Turkey’s public finances have been strongly influenced by a rather unstable economic
environment. As can be seen in Table 3, output growth has been very volatile in the first
half of this 5-year period, with sharp output contractions in 1999 and 2001 - reflecting the
negative impact of the earthquake in 1999 and the financial crisis in 2001 – and strong
- 21 -
rebounds after those crises. Overall, average economic growth has been low during this
period. By 2003, Turkey had just reached its pre-crisis output level. Inflationary pressures
have declined, although in 2001 the disinflation process had been interrupted by sharp
exchange rate depreciation after the switch from a crawling peg to a free floating exchange
rate. Nevertheless, annual consumer price inflation has declined from 65% in 1999 to 25%
in 2003. The sharp output contractions and low average growth led to a marked increase in
unemployment, rising from 6.5% of the labour force in 2000 to 10.5% in 2003. So far, the
fiscal impact of this deterioration in the labour market has been limited, mainly due to the
fact that the unemployment system has been introduced in 1999 only and the number of
entitled persons still is relatively small. Despite the sharp output fluctuations, the current
account remained relatively balanced, fluctuating between a deficit of 4.9% of GDP and a
surplus of 2.7%. So far, the external financing gap could be financed by short-term capital
inflows, which were attracted by the relatively high interest rate differential. However,
foreign direct investment inflows have remained negligible.
Table 3 - Turkey: Main economic trends
Growth
Inflation
Unemployment
Current account
Interest rate
Exchange rate
annual averages
1999
2000
2001
2002
2003
GDP in real terms, change in %
-4.7
64.9
7.7
-0.8
0.447
7.3
54.9
6.6
-5.0
0.575
-7.5
54.4
8.5
2.4
1.102
7.9
45.0
10.1
-0.8
1.436
5.8
25.3
10.5
-2.8
1.697
CPI, change in %
LFS, % of labour force
balance, % of GDP
5-year govt. bonds, % p. a.
TRL/EUR (million)
Source: Eurostat
Despite weak growth of disposable income and rising unemployment, the Turkish
authorities managed to achieve substantial primary surpluses. The main approach has been
to increase tax revenues by raising tax rates and to reduce public expenditures, especially by
moderate increases of public sector wages and linear expenditure cuts. However, education
and health have been largely exempted from those cuts.
3. Methodological issues
Main challenges
The data have been calculated by applying as far as possible the accounting and definitional
requirements (ESA 95) used for the fiscal surveillance of EU Member States. Some
progress has been made in this year’s notification in improving the accounting of accrued
interest and in providing data for the whole reporting period. The information on social
security institutions and local governments still is very limited.
Transposition of national budget balances into ESA 95
Table 4 summarises the transposition of the national budget balance into the ESA 95
general government net borrowing/net lending definition. The first line shows the actual
(and for 2004 planned) figures of the most prominent budget balance which is in Turkey the
central government budget deficit, which is the official budget announced by the Ministry
of Finance.
The transposition from the national budget concept into ESA 95 requires a series of
adjustments: First, the treatment of financial transactions - such as loan or equity sales and
- 22 -
Table 4 - Turkey:
Transposition of national budget balances into ESA 95 net lending (+)/borrowing (-)
2002
trillion
TRL
Most prominent national budget balance (2)
+ adjustment to central government net lending
= Central government net lending (S.1311)
+ Local government net lending (S.1313)
+ Social security net lending (S.1314)
= General government net lending (S.13)
2003
% of
GDP
-40,090 -14.5
2,515
trillion
TRL
2004 (1)
% of
GDP
-39,816 -11.0
trillion
TRL
% of
GDP
-46,399 -11.0
0.9
4,849
1.3
8,187
1.9
-37,575 -13.6
-34,967
-9.7
-38,212
-9.0
-0.1
1.0
-386
3,844
-0.1
1.1
-641
4,823
-0.2
1.1
-35,055 -12.7
-31,509
-8.7
-34,030
-8.0
-149
2,669
(1) planned
(2) Central government budget balance
purchases - has to be adjusted to ESA 95 standards. In the case of Turkey, the main change
is related to appropriations for guarantee payments and expected privatisation revenues.
However, the size of those corrections is very limited, reducing the deficit by 0.25% of
GDP in 2003 and 0.8% of GDP in 2004. Second, in another step, the accounting of accrued
receivables and payables such as revenues and interest payments is aligned with ESA 95
standards. In 2002, this adjustment reduced the central government deficit by 0.7% of GDP
and was rather neutral in 2003. For 2004, the Turkish authorities do not expect a significant
correction. A third adjustment refers to the institutional setup, taking into account the net
borrowing or lending of other central government bodies, which are not included in the
central government budget. In the case of Turkey, those are mainly extra-budgetary and
revolving funds. In 2003, those institutions registered a budgetary surplus of 0.8% of GDP,
which is expected to increase to 1.1% of GDP in 2004. And finally, adjustments for other
transactions, such as transfers related to the previous duty losses of state banks, are taken
into account, amounting to some 0.2% of GDP in 2003. Overall, these adjustments result in
a ESA 95 central government net lending, which during the reporting period was lower than
according to the national definition. In 2002, the ESA 95 central government net lending
was 0.9% of GDP lower than the national definition, in 2003 the ESA 95 figures were 1.3%
of GDP lower and for 2004 the deficit reducing impact is expected to be 1.9%.
As far as the other sectors of the general government are concerned, the main impact on the
overall balance comes from the social security institutions, reducing the general government
net lending by about 1% of GDP during 2002-2004. Local governments have only a very
limited impact on the ESA 95 general government deficit.
Stability of data6
As compared to the April 2003 notification there have been revisions of data, especially for
general government deficit for the years 2001-2003, mostly due to revision of central
government sub-sector.
The general government deficit worsened by 2.9 percentage points in 2001 (from 26.9% to
29.8%), and by 2.7 percentage points in 2002 (from 10.0% to 12.7%). The revisions for
2001 and 2000 are mostly due to a correction of the accrued interest on external debt which
had been previously misinterpreted, as well as a revision of accounts receivable related to
securities issued for the rehabilitation of some banks.
6
The remainder of this section has been provided by Eurostat.
- 23 -
For 2003, from a forecast of 11.3% of GDP, general government deficit improved to 8.7%,
mostly due to a better than expected budgetary outturn reported for central government, and
revised accruals calculations.
However, expenditure on interest of the general government sector remained unchanged and
seems very high in Turkey: 13.9% of GDP in 2000, 27.1% in 2001, 19.8% in 2002 and
18.3% in 2003. This may give rise to some further discussion with Eurostat.
For the past notifications, the Turkish authorities considered the coverage of the debt as
complete, given that only central government was borrowing funds. However, for the 2004
notification, when debt figures are provided by sub-sectors one can observe that also local
government and social security funds reported small amounts of debt.
Deficit and debt methodology
In the Turkish notification, the transition from deficit in public accounts to the net
borrowing of general government starts from the official central government budget
balance. Financial transactions have been excluded from the figures comprising this
balance. It seems that for previous years there have been no loans granted or repaid included
in the budget figures. However, in 2003 expenditure of the risk account (which is used for
payments on guaranteed debt) was recorded as a financial transaction and therefore
eliminated in the calculation of ESA 95 net lending/net borrowing.
Other accounts receivable and payable of central government cover mainly taxes. The
figures for accounts receivable seem very important for the whole period (56.4% of the
central government deficit for 2000, 14.6% for 2001, 27.6% for 2002 and about 41.8% for
2003).
Social contributions and social benefits are still recorded on a cash basis. However, accruals
for these can be expected for future notifications.
Accrued interest seems to have been correctly calculated, following improvement in the
recording of domestic debt transactions in the framework of a World Bank project.
External debt interest payments are calculated on an accrual basis using a simple method.
Net borrowing of the central government sub-sector is highly predominant within the
general government total. This sub-sector includes also the balancing item of extrabudgetary funds such as Privatisation Fund, Social Aid and Solidarity Incentive Fund,
Defence Industry Support Fund and Support Price Stabilisation Fund. Central government
sub-sector also contains the following agencies with their own special budget: TRT-General
Directorate of Turkish Radio and Television Institution, General Directorate of National
Lottery Administration, YURTKUR (General Directorate of Higher Education Credit and
Hostels Institution), and AOÇ (Atatürk State Farm). The inclusion of these in the general
government sector may be subject of further discussion with Eurostat.
Net lending/net borrowing for all the sub-sectors of general government were included in
the notification, but for local government and social security funds it is not sure that
information is complete.
For the compilation of the local government sub-sector an annual survey is conducted
providing figures on a cash basis. For the deficit and debt notification the results of the
survey are compared and completed with figures provided by central institutions like
Ministry of Finance, the Undersecretariat of Treasury, and social security institutions.
- 24 -
For the social security funds sub-sector their financial balance is used. However the figures
are available only on a cash basis for the moment.
In year 2001 a huge capital transfer to cover duty loss of public banks increased the central
government deficit to about 30 % of GDP. The nominal value for the debt has been used.
Central government holds debt of the local government, but the amounts are very low: 592
trillion TL in 2000, 1494 trillion TL in 2001, 1964 trillion TL in 2002 and 1832 trillion TL
in 2003.
In 2003 a depreciation of general government consolidated debt in foreign currency can be
noted, contrary to the situation in previous years when an appreciation was recorded.
There are still high statistical discrepancies between the deficit and the change in general
consolidated debt, of about 14.8% of the change in debt for 2001, 9.4% in 2002 and 19,5%
in 2003. An improvement is expected starting from the fiscal year 2004, due to some
progress being made in the budget classification in parallel with introducing the IMF's GFS
2001 classification.
Gross domestic product
Concerning the quality of GDP data, ESA 95 is adopted as a general framework. To
achieve a conceptual and material compliance numerous development projects have been
launched, including the exhaustiveness of national accounts data. Only after these projects
have been completed can an assessment be made as to full compliance.
4. Conclusions
Turkey’s public finances reflect the dramatic fiscal burden of the 2001 financial crisis and
illustrate the ongoing attempts to reduce the resulting fiscal imbalances. Despite a difficult
economic environment, the Turkish authorities achieved substantial primary surpluses after
2001. Those surpluses not only reduced the debt ratio, but also helped to bring down
interest rates by strengthening market confidence. The decomposition of Turkey’s debt
dynamics shows an increasing importance of maintaining significant primary surpluses and
of reducing the financing costs. While in previous years, high nominal GDP growth
strongly facilitated debt sustainability, the success in bringing down inflation reduces the
impact of nominal growth and increases the importance of other factors driving the
dynamics debt sustainability, such as the primary surplus or the differential between real
growth and implicit real interest rates.
Compared to the previous notification, there have been significant revisions of some
figures, mainly due to the status of data and to corrections for the misinterpretation of some
adjustments. It seems that the most important events have been correctly recorded.
However, there are some conceptual issues which should be clarified with Eurostat,
especially as far as delimitation of general government sector and accruals recording are
concerned. The National Statistical Office should be more involved in the preparation of
the notification figures, especially for the ESA 95 methodological aspects. A full
assessment of the data by Eurostat remains difficult. For further improvement of the quality
of figures, compilation of a full set of ESA 95 accounts (non-financial and financial) for the
general government sector and its sub-sectors is highly recommended by Eurostat.
- 25 -