Does Export Concentration Cause Volatility? Christian Busch 14. Januar 2010
Overview Countries with undiversified export structure are plausibly more vulnerable to external shocks. But difficult to evaluate causal influence of export concentration on external volatility (exchange rate, terms of trade, export growth) Use of a new instrument that is based entirely on geographic characteristics, and thus plausibly uncorrelated with other determinants, including institutions, income, and policies. Causal effect found for terms of trade and export growth, but not for exchange rates. 14. Januar 2010 2
Introduction Large variation in volatility of macroeconomic aggregates (output, exports, exchange rates, terms of trade etc.) across countries Volatility decreases with level of development Lack of access to financial markets to mitigate external risks External shocks can be amplified by weak institutional environment (Rodrik, 1999, JEG) Instability in economic performance critical for many developing countries Uncertainty lowers private investment and leads to slower growth Increases inequality and poverty Instability can result in crises 14. Januar 2010 3
Determinants of Volatility Policy Mismanagement (Fatas and Mihov, 2003, QJE) Financial Institutions (Easterly et al. 2001, WB) Political Institutions (Acemoglu et al., 2003, JME) Export Concentration and Geography (Malik and Temple, 2009, JDE) 14. Januar 2010 4
Determinants of Volatility: Institutional Constraints and Export Concentration Volatility of GDP per capita growth rate 0.05.1.15 GNB STP GEOROM HTI RWA TCD TJK LVA TZA AGO AZE BLR GAB UGA GUY BDI KAZ UZB LBN TTO MW I LCA COG ETH YEM NGA CAF UKR MRT ZAR BGR MOZ PER ARG SYC PNG SYR VCT ALB CMR KGZ RUS KNA SVK TGO KHM SLE IRN MLI MARATG URY MKD BRB DMA COM CUB JOR VNM ARM POL NIC CHL FJI NER ZWE MDA LTU SVN LSO SEN HKG CHN IDN BEN GMB KOR NAM THA PAN BFA CPV CYP CRI CZE BWA BRA ECU DOM GHA PRY PHL SGP ZMB GRDHUN MEX TUR BLZ SLV IRL DZA HND VEN EST FIN JAM MYS CAN ESP LUX ISL GIN BOL HRV CIV KENPL PRT TUN LKA MDG MLTGRC AUS BGD MUS NZL PAK CHE SWE GTM ZAF GBR USA COL DNK FRA INDNOR ISRJPN SWZ AUT DEU NLD ITA EGY BEL Volatility of GDP per capita growth rate 0.05.1.15.2 STP GNB GEO MRT ROM LVA SYR TJK BLR AZE HTI RWA GAB COGAGO BDI SGP LBN TZA NGA KAZ LCA CPV CYP MOZ GHA TGO GUY UKR UGA IRN DZA YEM BGR ZWE JOR TTOVCT MWI GMB SVK ALB NAM BWA DMA RUS LSO ARG KGZ NIC ATG BRB CMR PNG SYC KHM CAF COM MKD MUS KNA SLE MLI POL KEN MAR CHL PER NER ARM ETH HKG ZMB CHN SVN LTU MDA CIVFJI CUB VNM DOM ECU CZE BRA GRC ISR URY SEN EGYJAM BGD BFA CRI SLV HUN ISL JPN KOR PAN PRT IDN BOL BLZ BEN ESTFIN LUX MAC HNDVEN ESP TURMEX THA GRD GIN IND IRL NZL TUN PRY NPL ITA DNKCHE HRV CAN MYS PHL PAK MDG FRA BEL AUT AUS GTM MLT DEU NLD GBR USA SWE ZAF LKA NOR COL SWZ 0.2.4.6.8 Political constraints 1 2 3 4 Export concentration 14. Januar 2010 5
Export Concentration versus Income per Capita Export Concentration 1 2 3 4 BMU GNB NGA STP AGO BDI COM MRT COG GAB ETH MLI YEM RWA DZA BEN SLECAF VNM GIN TJKLSO TKM SYC QAT BTN GUY DMA NPL SDN PNG CPV BW A SAU KHM DJI ECU ISL MWI NER ERI BFA HTI GRD IRN PRY BGD BOL BLZ CUB OMN KW T UGA TGO GHA GMB MNG CMR SYR KNA BHR MOZ ZMB LCA VCT VEN MUS MDG DOM JAM PAN HND NIC TTO AZE EGY FJI KAZ CIV ARM MLT BHS TZA SEN PAK LKA NAM MAC ALB SLV PER COL CRI CHL HUN MDA KGZMAR SWZ BRB KEN IDN PHLGTM GEO TUN JOR URY ZW E RUS ATG MYS CYP NOR ISR MKD POL ARG LUX THA NZL LVA IND ROM LBN GRC IRL AUS MEX UKRHRV KOR FINCAN BRA LTU PRT SGP TUR JPN BLR EST BGR ZAF HKG CHN SVK SVN SWECHE DNK ESP AUT BEL GBR USA CZE DEU NLD ITA FRA Volatility of GDP per capita growth rate 0.05.1.15 GNB STP HTI RWA ROM GEO TCDTJK LVA TZA AGO AZE BLR UGA GUY GAB BDI UZB LBN TTO KAZ MWI LCA CAF COG ETH YEM NGA UKR MRT ZAR BGR MOZ PNG PER ARG ALB SYR VCT KHM CMR KGZ RUS KNA SYCSVK TGO MLI SLE MAR IRN MKD URY ATG VNM DMA BRB COMNIC ARM POL FJI NER MDAZW E JOR CUB LTU CHL PRI SVN LSO SEN BEN GMB NAM THA BFA CHN CPV IDN CRI PAN KOR DOM CYP BWABRA CZE HKG MAC ZMB PHL GHA SLV ECU GRD PRYTUR MEXHUN IRL HND BLZ VEN SGP DZA EST FIN JAM MYS ESP ISL CAN LUX NPL KEN GIN LKA BOL HRV PRT MDG CIV TUN GRC MLT AUS BGD MUS NZL PAK SWE IND GTM COL ZAF ISR GBRDNK ITA FRA AUT DEU BEL JPN NOR CHE USA SWZ NLD EGY 6 7 8 9 10 GDP per Capita 6 7 8 9 10 Income per capita 14. Januar 2010 6
Causality and Model Uncertainty Instrumental Variables Acemoglu et al. (2003, JME): Institutions versus policies Bayesian Methods Malik and Temple (2009, JDE): Geography and export concentration Panel Data Yang (2008, JM): Influence of democracy depends on ethnic homogeneity Here: New instrument for export concentration based on geographic characteristics 14. Januar 2010 7
Approach Use geographically determined component of export concentration as an instrument for actually observed export concentration to study influence on external volatility Remoteness as a natural barrier to trade (Malik and Temple, 2009) Transport costs affect sectors differently (imported intermediate goods) countries may be locked into concentrated export structures Export concentration determines how external shocks are absorbed Terms of trade: Undiversified countries are hit more severely by price shocks Exports: Diversification dampens exposure to demand shocks Exchange rates: currency demand fluctuates with exports 14. Januar 2010 8
Geography-based Instrument for Export Concentration Appropriately weighted average of countries remoteness (based on Frankel and Romer, 1999) Include information about distance, population, area, landlocked, border Do not use common geographic variables Distance from equator and tropical area (Sachs, 2003) Soil quality (Sokoloff and Engerman, 2000) Coastal area (Acemoglu et al., 2005) Control for trade Remoteness reduces trade; Increased specialization induces more volatility (di Giovanni and Levchenko, RES, forthcoming) 14. Januar 2010 9
New Zealand: Influence of Distance on Export Concentration export concentration 1.5 2 2.5 3 3.5 4 Nepal Guatemala Bolivia Venezuela Iran Peru Iraq Panama USSR Jordan Romania Jamaica Nigeria Trinidad Bangladesh Costa Rica Barbados and Tobago Syria Turkey Egypt Ecuador Bulgaria Greece Malta Sri Ghana Cyprus Mauritius Mexico Seychelles Lanka Oman Poland Myanmar Zimbabwe (Burma) Belgium Colombia Pakistan Kuwait Portugal Tanzania Qatar Saudi Arabia Vietnam Brazil Iceland Kenya Hungary Italy Uruguay Bahrain Austria Argentina Chile China India Germany Spain West United Arab Emirates France Philippines Korea South United Denmark Ireland FinlandKingdom Indonesia Thailand Canada Israel Norway Malaysia South Africa Sweden Switzerland Japan Netherlands United States Hong Kong Tuvalu Singapore New Caledonia Papua New Guinea Solomon Islands Kiribati Niue Fiji Cook Islands Australia 7 7.5 8 8.5 9 9.5 log distance 14. Januar 2010 10
Belgium: Influence of Trading Partner s Size on Export Concentration (Controlling for Distance) e( export concentration X ) -1 -.5 0.5 1 Niger Bahamas Guinea-Bissau Mongolia Myanmar (Burma) Chad Iraq Israel Central African Republic Malawi Liberia Sri Lanka India Nepal Nicaragua Seychelles Djibouti ComorosMauritania Panama Sierra Leone Haiti Benin Mozambique Guyana Jamaica Togo Suriname BarbadosGambia Fiji El Salvador Switzerland Malta Ireland Zambia Yemen Mali Syria Hungary Sudan Ethiopia Bahrain Trinidad Oman and Mauritius Tobago Paraguay Honduras Uganda Yugoslavia Bolivia Burkina Faso Bangladesh Gabon Congo Republic Iceland Jordan Dominican Guatemala Senegal Zimbabwe Republic Costa Rica Guinea Tunisia Cameroon United Kingdom Bulgaria Angola Netherlands Iran Qatar Cyprus Kuwait Norway Denmark Austria Hong Kong TanzaniaPakistan Ecuador Ivory Coast United Arab Emirates Sweden Madagascar Algeria Ghana Venezuela Romania Poland Thailand Greece Kenya Egypt Turkey Nigeria CanadaMexico Uruguay Finland Burundi Saudi Rwanda Portugal Morocco Peru Arabia Colombia SpainItaly Japan Congo Malaysia Democratic Philippines France Argentina Korea South Republic Brazil United States Indonesia Singapore New Zealand South Africa Chile Australia China -4-2 0 2 4 6 e( lpop2 X ) coef = -.08958648, se =.01800051, t = -4.98 14. Januar 2010 11
Method Estimate bilateral export concentration equation Include only purely geographic determinants Aggregate fitted values to estimate a geographic component of countries overall export concentration Problem: bilateral export concentration cannot be easily aggregated across countries (as is the case for trade) Solution: use an additively decomposable inequality measure 14. Januar 2010 12
Generalized Entropy Inequality Measure x value of trade in product h between country i and country j n number of products traded µ mean value of trade flows across products α inequality weighting parameter (set to 0.5) 14. Januar 2010 13
Concentration Measures Compared Herfindahl concentration 0.2.4.6.8 1 1 2 3 4 Generalized entropy concentration 14. Januar 2010 14
allows Decomposition by Export Country Where the first term is the concentration between all bilateral trade relationships for country i, and the second term is the concentration within each bilateral trade relationship Thus, we need to estimate X i,j, N i,j, and GE i,j 14. Januar 2010 15
Data Export concentration: Concentration across products (4-digit SITC) Comtrade dataset; averaged over 1980-2000 Volatility (standard deviation; WDI) Terms of trade Real effective exchange rate Export growth Gravity variables from Frankel and Romer (1999) Trade openness (EX+IM / GDP; WDI) Institutional environment (Aggregate Governance Indicators) 14. Januar 2010 16
How does Geography Influence Concentration? Dependent Variable Bilateral export concentration Bilateral exports Number of sectors ln Export ln Number Frankel/Romer all > 50 value of sectors original values Comtrade values ln distance 0.213*** 0.212*** -0.583*** -0.243*** -0.85*** -1.225*** (0.007) (0.008) (0.033) (0.011) (0.04) (0.036) ln population (country i) -0.262*** -0.246*** 0.663*** 0.270*** -0.24*** -0.411*** (0.003) (0.005) (0.020) (0.006) (0.03) (0.030) ln area (country i) 0.100*** 0.094*** -0.164*** -0.094*** -0.12*** -0.012 (0.003) (0.003) (0.016) (0.005) (0.02) (0.023) ln population (country j) -0.042*** -0.025*** 0.718*** 0.085*** 0.61*** 1.400*** (0.004) (0.005) (0.019) (0.007) (0.03) (0.028) ln area (country j) 0.010*** -0.005-0.184*** -0.011** -0.19*** -0.307*** (0.003) (0.004) (0.016) (0.005) (0.02) (0.023) Landlocked dummy 0.107*** 0.033** -0.905*** -0.148*** -0.36*** -1.149*** (0.011) (0.014) (0.060) (0.021) (0.08) (0.103) Common border dummy-0.231*** -0.075** 0.426*** 0.123*** (0.039) (0.038) (0.156) (0.045) Bilateral trade flows Constant 4.800*** 4.275*** 7.103*** 3.120*** -6.38*** -7.725*** (0.068) (0.086) (0.351) (0.120) (0.42) (0.469) R 2 0.358 0.304 0.319 0.226 0.360 0.455 Observations 12'503 6'773 6'773 6'773 3'220 6'284 14. Januar 2010 17
Actual versus Constructed Export Concentration Actual Export Concentration 1 2 3 4 IND CHN NGA AGO GNB BDI SOM COM COG GAB MRT SLB ETH RWA GIN YEM DZA BEN SLE MLI LBR CAF SUR QAT SYC MMR BTN DMA GUY NPL SDN SAU PNG IRN ECU DJI ISL BGD BFA HTI MWINER KWT GRD PRY BHR BOL BLZ GHACMR SYR ARE OMN UGA MUS GMB MNG MOZ VEN TGO ZMB VCT LCA DOM JAM MDGTTO HND NIC PAN EGY FJI CIV MLT BHS PAK LKA TZA SEN HUN MAR SLV PER COL CHL BRB CRI IDNPHL KEN TUN GTM JOR URY ZWE NOR ISRMYS CYP POL ARG THA NZL ROM GRC IRL AUS MEX FIN CAN JPN KOR SGP TUR PRT BRA HKG BGR ZAF CHE SWE DNK ESP AUT BEL GBR ITA USA NLD FRA 2 2.5 3 3.5 4 Constructed Export Concentration 14. Januar 2010 18
External Volatility and Export Concentration Dependent Variable Regressions of Volatility Terms of Trade Real Effective Exchange Rate Export Growth OLS IV OLS IV OLS IV (1) (2) (3) (4) (5) (6) Export 0.068*** 0.043*** 0.054*** 0.006 0.076*** 0.054*** Concentration (0.008) (0.013) (0.012) (0.023) (0.010) (0.019) Trade -0.040*** -0.038*** -0.066*** -0.056*** -0.040*** -0.035*** Openness (0.009) (0.009) (0.019) (0.019) (0.012) (0.013) Constant 0.087** 0.143*** 0.248*** 0.326*** 0.121** 0.161*** (0.039) (0.055) (0.071) (0.098) (0.047) (0.058) R 2 0.404 0.353 0.323 0.139 0.324 0.296 N 110 110 73 73 127 127 Excluded Instrument Predicted Predicted Predicted Concentration Concentration Concentration Levels of statistical significance indicated by: *** p<0.01; ** p<0.05; * p<0.1 Numbers reported in parentheses are heteroscedasticity-robust standard errors. 14. Januar 2010 19
Importance of Controlling for Trade (e.g. ToT) -1.5-1 -.5 0.5 1 Residuals from Regressing Exp. Conc. on Pred. Trade -2-1 0 1 2 Residuals from Regressing Trade on Pred. Exp. Conc. -1 -.5 0.5 1 Resdiuals from Regressing Predicted Export Concentration on Predicted Trade -2 0 2 4 6 Resdiuals from Regressing Predicted Trade on Predicted Export Concentration -1 -.5 0.5 1 1.5 Residuals from Regressing Exp. Conc. on Pred. Exp.Con. -3-2 -1 0 1 2 Residuals from RegressingTrade on Pred. Trade -2 0 2 4 6 Resdiuals from Regressing Predicted Trade on Predicted Export Concentration -1 -.5 0.5 1 Resdiuals from Regressing Predicted Export Concentration on Predicted Trade 14. Januar 2010 20
Export Concentration and Institutions Regressions of Volatility in Terms of Trade Method OLS OLS IV IV IV 2SLS First Stage First Stage Exp. Conc. Governance (1) (2) (3) (4) (5) Export 0.046*** 0.047*** 0.042* Concentration (0.010) (0.014) (0.025) Trade -0.017** -0.011-0.011 0.008 1.130 Openness (0.008) (0.008) (0.007) (0.107) (0.928) Aggregate -0.005*** -0.004** -0.003 Governance (0.001) (0.002) (0.004) Predicted Exp. 0.505** 2.540 Concentration (0.230) (2.067) Settler 0.323*** -2.394*** Mortality (0.044) (0.458) Constant 0.057* 0.024 0.042-0.302-1.394 (0.034) (0.036) (0.067) (0.964) (6.893) R 2 0.468 0.384 0.374 0.512 0.431 N 104 58 58 58 58 Excluded Instrument Predicted Predicted Predicted Concentration, Concentration, Concentration, Settler Mortality Settler Mortality Settler Mortality Levels of statistical significance indicated by: *** p<0.01; ** p<0.05; * p<0.1 Numbers reported in parentheses are heteroscedasticity-robust standard errors. 14. Januar 2010 21
Robustness Checks Overidentification tests using concentration across countries (predicted from weighted bilateral distance) as additional IV Control for export shares of various natural resources (oil, agriculture, various raw materials) Control for the size of the manufacturing sector: inconclusive results 14. Januar 2010 22
Conclusions Bilateral export concentration is well described by a gravity equation Inequality decomposition methods allow to aggregate bilateral export concentration to an overall measure of export concentration Predicted export concentration contains valuable information about actual concentration Using predicted export concentration as an IV, export concentration exhibits a causal influence on volatility in terms of trade and growth in exports, but not so on volatility in exchange rates The instrumental variable is not suited to elucidate the separate roles of institutions and export concentration in determining aggregate volatility 14. Januar 2010 23