Chapter

Appendix I Data and Definitions

Author(s):
Mark Horton, George Tsibouris, Wojciech Maliszewski, and Mark Flanagan
Published Date:
June 2006
Share
  • ShareShare
Show Summary Details

An adjustment episode is defined as any year or succession of years of uninterrupted improvement in the fiscal primary balance, and focuses on the consolidated central government (on a cash basis). In setting this definition, a number of considerations are taken into account:

  • The primary balance is a more appropriate indicator of the fiscal stance, given the volatility of interest and exchange rates. Indeed, during adjustments accompanied by disinflation, excluding interest is vital to avoid mismeasurement of fiscal effort: in these circumstances, the interest bill can include a large debt amortization component.
  • The focus on uninterrupted improvement avoids double counting. Earlier studies that were based on a predetermined duration for an adjustment episode often considered contiguous episodes as separate consolidations, potentially biasing results.
  • Noncyclically adjusted data were used due to data limitations. Cyclical adjustment may not be of critical importance in this type of analysis. For example, Adam and Bevan (2004) found no relation between lagged GDP and the fiscal balance in their sample of 127 countries.
  • Data for the consolidated central government on a cash basis were used also due to data limitations. Using data on a commitment basis or using a general government or even broader public sector concept (and including contingent liabilities such as pension entitlements) would have better reflected the underlying fiscal stance, but such data were not available for the full sample of countries.
  • Data on fiscal outcomes were used, again due to data limitations. To limit cases of unintended adjustment, episodes for oil exporters were excluded for periods of significant increases in real oil prices.18 Cases of targeted “large” adjustments that do not reach the threshold would also present significant conceptual problems, because targets can change over the course of an adjustment.
Table A1.1.Data Set: Country Coverage1
Developed Countries (24)Emerging Markets (22)Transition Economies (24)Developing Economies (95)
AustraliaArgentinaAlbaniaAlgeriaEcuadorLibyaSamoa
AustriaBrazilArmeniaAntigua and BarbudaEl SalvadorMadagascarSâo Tomé and Principe
BelgiumChileAzerbaijanBahamas, TheEquatorial GuineaMalawiSaudi Arabia
CanadaColombiaBelarusBahrainEritreaMaldivesSenegal
DenmarkCyprusBulgariaBangladeshEthiopiaMaliSeychelles
FinlandEgyptCroatiaBarbadosFijiMauritaniaSierra Leone
FranceIndiaCzech RepublicBelizeGabonMauritiusSolomon Islands
GermanyIndonesiaEstoniaBeninGambia, TheMoroccoSri Lanka
GreeceIsraelGeorgiaBhutanGhanaMozambiqueSuriname
IcelandJordanHungaryBoliviaGrenadaMyanmarSwaziland
IrelandKorea, Rep. ofKazakhstanBotswanaGuatemalaNamibiaSyrian Arab Republic
ItalyMalaysiaKyrgyz RepublicBurkina FasoGuineaNepalTanzania
JapanMexicoLatviaBurundiGuinea-BissauNetherlands AntillesTogo
LuxembourgPakistanLithuaniaCambodiaGuyanaNicaraguaTonga
MaltaPeruMacedonia, FYRCameroonHaitiNigerTrinidad and Tobago
NetherlandsPhilippinesMoldovaCape VerdeHondurasOmanTunisia
New ZealandSingaporeMongoliaCentral African Rep.Iran, I.R. ofPanamaUganda
NorwaySouth AfricaPolandChadJamaicaPapua New GuineaVanuatu
PortugalThailandRomaniaComorosKenyaParaguayYemen, Rep. of
SpainTurkeyRussiaCongo, Rep. ofKiribatiQatarZambia
SwedenUruguaySlovak RepublicCosta RicaKuwaitRwandaZimbabwe
SwitzerlandVenezuelaSloveniaCôte d’IvoireLao P.D.R.St. Kitts and Nevis
United KingdomTajikistanDjiboutiLebanonSt. Lucia
United StatesUzbekistanDominicaLesothoSt. Vincent and the Grenadines
Dominican Rep.Liberia
Sources: IMF, Government Finance Statistics (various issues), international Financial Statistics (various issues), World Economic Outlook, and Monitoring of Fund Arrangements; IMF staff estimates; and country authorities.

Countries excluded due to lack of data are Angola, Brunei Darussalam, China, Democratic Republic of the Congo, Hong Kong SAR, Nigeria, Sudan, Taiwan Province of China, Turkmenistan, and Ukraine.

Sources: IMF, Government Finance Statistics (various issues), international Financial Statistics (various issues), World Economic Outlook, and Monitoring of Fund Arrangements; IMF staff estimates; and country authorities.

Countries excluded due to lack of data are Angola, Brunei Darussalam, China, Democratic Republic of the Congo, Hong Kong SAR, Nigeria, Sudan, Taiwan Province of China, Turkmenistan, and Ukraine.

Table A1.2.Data Sources and Transformations
Variable ListSources
Fiscal (at the central government level)For all fiscal and financing data: IMF, Government Finance Statistics (GFS); country authorities; and IMF staff estimates.
Total revenues and grants
Tax revenues, of which:
Taxes on income, profits, and capital gains
Domestic taxes on goods and services
Taxes on international trade and transactions
Nontax revenues
Capital revenues
Grants
Total primary expenditure
Primary current expenditure, of which:
Subsidies
Goods and services (excluding wages)
Wages
Capital
Net lending
Domestic financing
External financing
Macroeconomic
Real GDP growthIMF, World Economic Outlook; and country authorities.
CPI inflationIMF, International Financial Statistics (IFS), World Economic Outlook; country authorities; and IMF staff estimates.
Trade balanceIMF, IFS, and World Economic Outlook.
Real consumption growth (deflated by CPI)IMF, IFS; and country authorities.
Real investment growth (deflated by CPI)IMF, IFS, and World Economic Outlook.
Central government debtIMF, GFS; country authorities; and IMF staff estimates.
Other
IMF program years by countryIMF, Monitoring of Fund Arrangements.
Sources: IMF, Government Finance Statistics (various issues), International Financial Statistics (various issues), World Economic Outlook, and Monitoring of Fund Arrangements; IMF staff estimates; and country authorities.
Sources: IMF, Government Finance Statistics (various issues), International Financial Statistics (various issues), World Economic Outlook, and Monitoring of Fund Arrangements; IMF staff estimates; and country authorities.
18Specifically, for countries with oil exports in excess of 25 percent of GDP during any five-year period, an episode was excluded if it took place during 1973, 1979–80, 1999–2000, when real oil prices increased by over 10 percent. This eliminated 35 episodes, including 20 that would be considered large. Use of the nonresource primary balance would best address the exogeneity problem. The data necessary to construct these balances are not readily available, however.

    Other Resources Citing This Publication