Global Recessions

This will help you further understand how financial crises lead to economic recession and examine this notion from a global perspective by further examining the 2008 financial crisis. What is a depression in the context of financial markets?

Multiple data sources are employed to construct world GDP growth at annual and quarterly frequencies over a long period. The annual GDP series covers 180 countries –  36 advanced economies and 144 EMDEs – over the period 1950-2019, though the country sample size varies by year.  The quarterly series covers 106 countries over the period 1960:1-2019:3.  Appendix 1 presents the list of countries used to compute global growth. In addition to data on GDP growth, a wide range of measures of global economic and financial activity are employed.  Indicators of economic activity include trade, industrial production, unemployment, and oil consumption. Financial variables include capital flows, credit, equity and house prices, inflation, short-term nominal and real interest rates, and an index of broad financial conditions. In light of their roles in determining activity, some additional series, such as indicators of uncertainty and confidence, are also examined. Appendix 2 includes a summary of all variables in the database with their definitions, coverage, and sources. 

Appendix 1. List of countries in the database

Advanced economies (36)     
Australia Hong Kong SAR„ China  New Zealand 
Austria Iceland Norway 
Belgium  Ireland  Portugal 
Canada Israel Singapore 
Cyprus Italy Slovak Republic 
Czech Republic  Japan  Slovenia
Denmark  Korea, Rep. Spain
Estonia Latvia  Sweden 
Finland  Lithuania Switzerland 
France Luxembourg Taiwan, China 
Germany  Malta  United Kingdom 
Greece Netherlands United States
Emerging market and developing economies (144)  
Cambodia* Micronesia, Fed. Sts.*  Thailand
China Mongolia Timor-Leste* Tonga*
Fiji* Myanmar* Tuvalu*
Indonesia Palau* Vanuatu*
Kiribati* Papua New Guinea*  Vietnam
Lao PDR* Philippines
Malaysia Samoa*
Marshall Islands* Solomon Islands*
Europe and Central Asia (24)     
Albania Hungary  Romania
Armenia*  Kazakhstan  Russian Federation
Azerbaijan Kosovo*  Serbia
Belarus Kyrgyz Republic*  Tajikistan*
Bosnia and Herzegovina  Moldova  Turkey
Bulgaria Montenegro  Turkmenistan*
Croatia  North Macedonia  Ukraine
Georgia Poland  Uzbekistan*
Latin America and the Caribbean (30)  
Antigua and Barbuda*  Dominican Republic  Panama
Argentina  Ecuador  Paraguay
Bahamas, The*  El Salvador  Peru
Barbados*  Grenada*  St. Kitts and Nevis*
Belize  Guatemala  St. Lucia*
Bolivia  Guyana* St. Vincent and the Grenadines*
Brazil  Haiti*  Suriname*
Chile  Honduras  Uruguay
Colombia  Jamaica*
Costa Rica  Mexico
Dominica*  Nicaragua
Middle East and North Africa (16)    
Algeria*  Jordan  Saudi Arabia
Bahrain  Kuwait  Tunisia
Djibouti*  Lebanon*  United Arab Emirates
Egypt, Arab Rep.  Morocco  West Bank and Gaza
Iran, Islamic Rep.  Oman*
Iraq*  Qatar
Sub-Saharan Africa (44)    
Angola*  Ethiopia*  Namibia
Benin*  Gabon*  Niger*
Botswana  Gambia, The*  Nigeria
Burkina Faso*  Ghana  Rwanda*
Burundi*  Guinea*  Senegal
Cabo Verde*  Guinea-Bissau*  Seychelles*
Cameroon  Kenya  Sierra Leone*
Chad*  Lesotho  South Africa
Comoros*  Liberia*  Sudan*
Congo, Dem. Rep.*  Madagascar*  Tanzania
Congo, Rep.*  Malawi*  Togo*
Côte d’Ivoire  Mali*  Uganda
Equatorial Guinea*  Mauritania*  Zambia
Eritrea*  Mauritius*  Zimbabwe*
Eswatini*  Mozambique

Appendix 2. Definitions and sources of variables

Variable  Definition  Source
Economic activity
Output GDP in constant 2010 U.S. dollars (market weighted), taken from the World Bank. Prior to 1960, the series is extended using data from the Penn World Tables 9.1 (PWT 9.1). Sample includes 180 countries, including 36 advanced economies and 144 EMDEs. For PPP-weighted series, GDP is in constant 2010 international dollars. Feenstra, Inklaar, and Timmer (2015); World
Bank.
Output per capita GDP per capita, in constant 2010 U.S. dollars (market weighted), taken from the World Bank. Prior to 1960, both GDP and population series are extended using data from PWT 9.1. Sample includes 180 countries. For PPP-weighted series, GDP per capita is in constant 2010 international dollars. Feenstra, Inklaar, and Timmer (2015); World Bank.
Industrial production Industrial production index (if not available, manufacturing production index is used). Data are obtained at a quarterly frequency and used as annual averages of year-on-year growth rates. For aggregated groups, market-weighted output is used as a weight. The main source of the series is the World Bank and, for countries and quarters without data, the series is extended using growth rates of data from the Organisation for Economic Co-operation and Development OECD) and Haver Analytics. Haver Analytics, OECD, World Bank.
Trade Exports plus imports of goods and services, in constant 2010 U.S. dollars, taken from the World Bank. Prior to 1960, the series of exports and imports are extended using data from PWT 9.1. Trade for aggregated groups is the sum of exports and imports of individual countries. Feenstra, Inklaar, and Timmer (2015); World Bank.
Unemployment rate Unemployment, in percent of labor force, taken from the International Monetary Fund (IMF). Data for aggregated groups are computed as the sum of those unemployed in individual countries divided by the sum of labor force. IMF.
Oil consumption Oil consumption, in thousand barrels per day. Oil consumption in advanced economies and EMDEs is computed as the sum of consumption in individual countries. If there are differences between the world total and the sum of advanced economies and EMDEs, the residuals are added to the EMDE aggregate. The number of world oil consumption for 2019 is based on data over the first nine months of the year. British Petroleum, U.S. Energy Information
Administration.
Financial market
Capital flows Total capital flows, defined as the sum of absolute values of outflows (net acquisition of financial asset, including direct, portfolio, and other assets) and inflows (net incurrence of liabilities, including direct, portfolio, and other liabilities), in current U.S. dollars, taken from the IMF Balance of Payments Statistics (sixth edition). The figures for 2019 show data over the first two quarters of the year. For the historical series, the statistics based on the fifth edition are also used. The series is shown as a percent of GDP, which is taken from the World Bank and PVT 9.1.  Feenstra, Inklaar, and Timmer (2015); IMF; World Bank.
Credit Nominal credit provided by banks and other financial corporations, deflated by the consumer price index (CPI). Data are at a quarterly frequency and shown as annual averages of year-on-year growth of real credit. Nominal credit series is taken from the IMF (International Financial Statistics: IFS), titled claims on private sector (by depository or financial corporations), and the Bank for International Settlements. CPI is taken from the Bank for International Settlements, Haver Analytics, and the IMF. Credit for aggregated groups is computed as follows. First, real credit (in local currency) is converted to constant 2010 U.S. dollars, and then U.S.-dollar real credit in individual countries is aggregated into respective groups.  Bank for International Settlements, Haver Analytics, IMF.
Equity prices Share price index, deflated by CPI. Data are at a quarterly frequency and used as annual averages of year-on-year growth of real equity prices. Data are from the IMF (IFS) and Haver Analytics and available as period averages and end-of-period values. The one with longer data availability is used as the main series. Growth in aggregated groups is computed with market-weighted output as a weight. CPI is taken from the Bank for International Settlements, Haver Analytics and the IMF.  Bank for International Settlements, Haver Analytics, IMF, World Bank.
House prices House (or property) prices, deflated by CPI. Data are at a quarterly frequency and used as annual averages of year-on-year growth of real house prices. Data are from the Bank for International Settlements, Haver Analytics, and OECD. Growth in aggregated groups is computed with market-weighted output as a weight. CPI is taken from the Bank for International Settlements, Haver Analytics, and the IMF.  Bank for International Settlements, Haver Analytics, IMF, OECD, World Bank.
Financial conditions An index of financial conditions, computed as a weighted average of short-term and long-term interest rates, trade-weighted dollar, an index of credit spreads, and the ratio of equity prices to the 10-year average of earnings per share, as explained in Hatzius and Stelin (2018). Higher index numbers reflect tighter financial conditions. Data are at a monthly frequency and used as ammual averages of year-onyear growth of the index. Growth in aggregated groups is computed with market-weighted output as a weight. Bloomberg, Goldman Sachs, World Bank.
Inflation  Change in CPI in percent. Data are at a quarterly frequency and used as annual averages of year-on-year percent changes of CPI. The series is taken from the Bank for International Settlements, Haver Analytics, and the IMF. Inflation in aggregated groups is computed with market-weighted output as a weight. In order to eliminate an effect of countries with a history of high inflation, countries whose average inflation over the entire sample period (based on quarterly data) is above the top 20th percentile, are excluded. Bank for International Settlements, Haver Analytics, IMF, World Bank.
Interest rates, confidence, and uncertainty
Shor-term interest rate (nominal and real) Treasury bill rates or money market rates, with the maturity of three months or less. In countries where 3-month rates are unavailable, shorter maturity rates (including overnight rates) are used. Data are at a quarterly frequency and used as annual averages. The real short-term interest rate is the difference between nominal rate and inflation in the following quarter (as a proxy of expected inflation). Data are taken from different sources, and the data series with the longest time coverage is used as the main series. Then, the main series is spliced with those from other data sources. Data for aggregated groups are computed with market-weighted output as a weight. Bank for International
Settlements, Haver
Analytics, IMF, OECD,
World Bank.
Business confidence Business confidence index, originally taken from the European Commission, OECD, and country sources (including statistical offices, central banks, academic institutions, and think tanks). The series is obtained as quarterly data and used as annual averages. Data from different sources are first standardized and converted to an index equal to 100 in 2015-18. Confidence for aggregated groups is computed with market-weighted output as a weight. European Commission,
OECD, country sources,
World Bank.
Policy uncertainty Economic policy uncertainty index, which is based on the frequency of articles in domestic newspapers mentioning economic policy uncertainty. Country-level indices are at a monthly frequency and used as annual averages. Data are first converted to a 6-month moving average and then an index equal to 100 in 2015-18. Aggregated policy uncertainty is computed with market-weighted output as a weight, based on 20 countries with available data. Davis (2016); World
Bank.
Policies
Government expenditure Government primary expenditure (i.e., government total expenditure excluding interest expense) in local currency, deflated by GDP deflator. The expenditure series is first taken from the IMF and then extended with data in Mauro et al. (2015). GDP deflator is taken from the IMF, PWT 9.1, and the World Bank. Real government expenditure growth in aggregated groups is computed with marketweighted output as a weight. IMF; Mauro et al. (2015); World Bank.
Policy interest rate Nominal central bank policy rates. Data are at a quarterly frequency and used as annual averages. Data are taken from different sources, and the data series with the longest time coverage is used as the main series. Then, the main series is spliced with those from other data sources. Data for aggregated groups are computed with market-weighted output as a weight. Bank for International
Settlements; Haver
Analytics; IMF; OECD; World Bank.
Quarterly series 
Output (per capita) growth Quarterly real GDP is taken from Haver Analytics and OECD. Quarter-on-quarter annualized growth rates of seasonally-adjusted real GDP are computed. If the original data are not seasonally-adjusted, the U.S. Census X-13 program is used to perform seasonal adjustment first. Quarterly growth for aggregated groups is computed with market-weighted annual output as a weight—the same weights are applied to four quarters in given years. In computing per capita growth, population data are used on an annual basis and taken from the World Bank and PWT, and therefore, the same population growth numbers are used in four quarters in given years. As a result, output per capita growth is calculated as the difference between aggregated annualized quarterly GDP growth and annual population growth. When computing growth rates over two quarters with different samples, countries are restricted to the common samples between these two quarters and with GDP and population. Sample includes 106 countries, though the sample size varies by quarter, over 1960:1-2019:3. Data for the 1950s are excluded, due to the limited data availability. Feenstra, Inklaar, and Timmer (2015); Haver Analytics; OECD; World Bank.


The main measure of the global business cycle is the growth rate of world real GDP per capita.  Real GDP per capita is considered as a primary measure of average economic well-being since it takes into account differences in population growth. The difference between per capita GDP growth rates in advanced economies and EMDEs is generally smaller than the difference between their aggregate GDP growth rates.  

The growth rate of world real GDP is a weighted average of national real GDP growth rates. Two types of weights are employed: market exchange rate weights and purchasingpower-parity (PPP) weights. The baseline results refer to market exchange rate weights, which are calculated as national GDP measured in domestic currencies, converted into U.S. dollar terms using market exchange rates, as a share of world GDP in U.S. dollar terms. Global trade and transactions in financial markets are conducted at market exchange rates, and the baseline specification uses this weighting scheme. 

PPP exchange rates are calculated as the rates at which the currency of one country would have to be converted into another to equalize the values of a common and broadly defined basket of goods and services. PPP exchange rates differ from market exchange rates particularly because goods and services that are not traded internationally tend to be cheaper in lower-income countries than in higher-income countries. As a result, the value of output in lower-income countries tends to be relatively greater using PPP than using market exchange rates. Thus, PPP weights, which are calculated as national GDP valued at PPP as a share of world GDP, tend to be higher for lower-income countries than market exchange rate weights. As growth in lower-income countries tends to be greater than that in higher-income countries, global GDP growth is often higher with PPP weights than with market weights.  

For measuring living standards and aggregating welfare, PPP weights are more appropriate since they capture the amount of consumption affordable to households for comparable consumption baskets. While PPP weights capture the fact that some goods are cheaper in lower income countries, market rates capture how much an economy could "buy" in global markets. Hence, market exchange rate-based weights are used here to provide the baseline measure of economic size.