## wpiea2020133-print-pdf - Section 8 concludes. ## Source details **Canonical URL:** [wpiea2020133-print-pdf - Section 8 concludes.](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020133-print-pdf.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020133-print-pdf.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020133-print-pdf.pdf.json) --- ### COVID-19 literature and the paper's contribution - Rapidly growing literature integrates infection dynamics into macroeconomic analysis, often via SIR models or extensions. - Representative findings from cited studies: - Stock (2020) and Alvarez et al. (2020): optimal policy is a full lockdown covering the majority of the population with restrictions removed gradually afterwards. - Acemoglu et al. (2020): multi-risk SIR model by age groups; targeted measures such as full lockdown for the elderly could be more effective. - Alon et al. (2020): developing-country perspective with market distortions and an informal sector; conclude full economy lockdowns are infeasible and lockdowns targeting the elderly might be better. - Farboodi et al. (2020): individuals sharply reduce activity in laissez-faire; socially optimal response imposes severe early restrictions. - Eichenbaum et al. (2020): incorporating supply and demand with government-altered activity via a consumption tax; find severe early containment is socially optimal. - Krueger et al. (2020): introduce differential transmission rates based on consumption or employment choice; highlight importance of sectoral heterogeneity. - Barrot et al. (2020), Bonadio et al. (2020), Baqaee et al. (2020): incorporate input-output (I-O) linkages to analyze propagation of sectoral supply shocks. - Baqaee and Farhi (2020b): Keynesian model with sectoral supply shocks and aggregate demand shocks; show large amplification in a network economy (absent infection dynamics). - Guerrieri et al. (2020): multi-sector economy where supply shocks can turn into larger aggregate demand shocks (no infection dynamics). - Baqaee and Farhi (2020a): in a non-SIR general equilibrium with nonlinear production networks, nonlinearities amplify the impact of COVID-19 between 20 to 100 percent; reductions in labor supply are less costly if spread across many sectors. - Distinctive contribution of this paper: - First paper (to the authors' knowledge) to incorporate both supply and demand shocks at the sector level for an open economy operating within the international production network. - Estimates COVID losses considering domestic and foreign sectoral shocks and their amplification through I-O linkages. - Links economic losses to I-O trade links and capital flows empirically to inform fiscal needs of emerging markets (EMs). ### Theoretical framework and lockdown scenarios - Model schematic elements: - SIR model with industrial heterogeneity (teleworkable vs on-site workers; physical proximity). - Supply channel: proportion of workforce impaired by infection, dependent on teleworkable share and physical proximity. - Demand channel: consumption changes linked to infection levels and mitigated by lockdowns. - Inter-industry and international linkages determine propagation and equilibrium output (minimum of supply- and demand-implied output). - Lockdown scenarios implemented: - Partial lockdown: - all industries remain open; - teleworkable portion of employees work from home; - low infection rate for teleworkables and the general public; - high infection rate for on-site workers. - Full lockdown: - only essential industries remain open; - workers in non-essential sectors stay at home; - infection rates lowered for almost everyone. - Key model calibration and identification: - Teleworkable employment share by industry from Dingel and Neiman (2020). - Physical proximity proxy from O*NET Work Context "Physical Proximity". - Recovery rate γ = 1/14 ≈ 0.07. - World Health Organization reported R_0 between 2 and 2.5; the lower end, R_0 = 2, is used. - In absence of industrial heterogeneity, R_0 = 2 and γ = 0.07 implies β = 0.14. - Imposed condition linking β_0 and Prox_i: β_0 = β ( 1 + sum_{i=1}^K Prox_i N_i / N )^{−1} with β = 0.14. - Production and supply: - Output linear in labor: Y_{i,t} = Z_i L_{i,t}. - Pandemic-period available workers: ˜L_{i,t} = (N_{i,t} − I_{i,t}) + TW_i (1 − I_{0,t} / N_0). - Pandemic-period sector output: Y^S_{i,t} = Z_i ˜L_{i,t}. - Equilibrium rule: - Final industry-level production is the element-wise minimum of supply-implied output and demand-implied output. ### Demand modeling, data, and mapping to I-O framework - Demand channels and representation: - Two demand profiles: normal times (national accounts consumption) and brunt of pandemic (credit card data and other real-time proxies). - Representative agent utility: Cobb-Douglas U(e1, ..., en) = ∏_{i=1}^n e_i^{α_i} with ∑_{i=1}^n α_i = 1. - During pandemic: ̃U with ̃α_i(I) changing with infections I; income reduced by factor 1−η(I). - Scaling parameter ̄I = 50,000; threshold I ≤ 0.1 ̄I corresponds to I ≤ 5,000. - Demand ratio: δ_i(I) = ̃α_i(I) η(I) / α_i with δ_i(I) = 1 for I ≤ 0.1 ̄I and ̄δ_i = ̄α_i ̄η / α_i as I → ∞. - Functional form for smooth transition specified: δ_i(I) = 1 if I ≤ 0.1 ̄I; δ_i(I) = ̄δ_i / [1 + (I/̄I − 0.1) ̄δ_i + (I/̄I − 0.1)] if I > 0.1 ̄I. - Mapping to final demand and output: - Revised final demand: ̃F_{c,i}(I_t) = F_{c,i} δ_i(I_t). - Use OECD Inter-Country Input-Output (ICIO) Tables (2015): 36 industries × 69 entities → 2484 × 2484 matrix. - A matrix: direct requirements; equilibrium Y = (I − A)^{-1} F. - Demand-driven output during infections: Y^D_t = (I − A)^{-1} ̃F(I_t). - Data sources and empirical proxies: - CBRT credit card data used for peak-pandemic demand proxies; other sectors use industry reports and historical data. - Aggregate demand shock magnitude: 23% when focusing only on sectors with credit card data; 16% when considering the full set of sectors. - Teleworkable share and proximity indices constructed from Dingel and Neiman (2020), O*NET, and BLS OES mapped to OECD ISIC codes. ### Open-economy implications and global coordination - Global synchronization of lockdowns: - If lockdowns are globally synchronized, pandemic is controlled faster. - Faster global decline in infections leads to quicker normalization of domestic and foreign demand and minimizes economic costs. - External amplification: - International I-O linkages and capital flows amplify domestic sectoral shocks through trade and external finance channels. - Equilibrium output uses international input-output structure to capture propagation. ### Initial conditions in Turkey and external financing needs - Pre-pandemic macro background and vulnerabilities: - Since 2017, inflation rose while Turkish Lira (TL) depreciated. - August 2018 marked an exchange rate crisis triggered by political tension with US; rapid TL depreciation pushed FX-indebted firms toward bankruptcy. - Growth rate in Q1 2020: 4.5 percent. - Unemployment rate declined to 12.7 percent. - As of first week of April 2020, net reserves of CBRT: $26 billion, of which $25 billion was borrowed from domestic banks. - IMF-defined budget deficit excluding one-time transfers: close to 5 percent of GDP (average over last 5 years). - Current account deficit: around 2.5 percent of GDP (average over last 5 years). - External debt and currency composition: - Turkey’s external debt: 60 percent of GDP at end-2019. - Historical context: 2001 total external debt: 57 percent of GDP (public sector 24 percent, private sector 22 percent). - By 2019, total external debt comparable to 2001 with 56 percent of GDP; private sector share 36 percent of GDP, public debt 21 percent of GDP. - As of December 2019, almost 60 percent of total external debt is denominated in USD. - Maturity and short-term rollover needs (verbatim figures preserved): - Total external debt: $437 billion. - Short-term portion: $124 billion (17 percent of GDP). - Of short-term debt, $93 billion held by private sector. - BIS data: $96 billion of total external debt belongs to the banking system. - External debt needing rollover in 2020: $169 billion (approximately 23 percent of GDP). - Banking sector’s share in short-term debt: $81 billion. - Rollover scenarios: - If rollover ratios stay at current levels, Turkey needs around $30 billion in 2020. - If rollover ratios fall to GFC levels, Turkey might need around $90 billion in 2020. - Market sentiment and capital flows (January–April 2020): - From beginning of year until week of April 24, 2020: - $2.7 billion of equity held by foreign investors sold to domestic investors. - $5.5 billion of government bonds held by foreign investors sold to domestic investors. - Corporate bond sell-off from January 3 to April 24, 2020: $86.5 million outflow. - Banking sector cross-border flows: - In January 2020, Turkish banks paid net $0.8 billion more in short-term loans than they borrowed, and $0.7 billion more in long-term loans. - In April 2020, Turkish banks paid $1 billion in long-term loans. - FX exposure amplifies COVID-19 economic costs: - More than 1/3 of total external funding is obtained through bank loans in Turkey, almost all in FX. - Half of entire corporate sector debt is in FX and most is borrowed from domestic banks. - Sectors with higher FX exposure experience larger output losses under the COVID-19 shock (correlation reported: corr: 0.558 & t-stat: 2.231). - Figure 5 (descriptive): plots 2016 sectoral FX exposure (foreign currency debt / total debt) against estimated sectoral output loss (%) under a no-lockdown, no-policy scenario; shows stronger FX funding → sharper output declines. - Policy implication from initial conditions: - High FX corporate exposure and sizable external debt imply that exchange rate depreciation acts as an amplification mechanism exacerbating bankruptcies and output losses in FX-exposed sectors. - Well-designed policies needed to address loan restructuring, non-performing loans, and non-viable firms given private-sector FX debt concentration. ### Monetary, financial, and fiscal policy responses (Turkey context) - Monetary and financial measures: - CBRT cut rates by 100 basis points immediately during their emergency meeting on March 18, 2020 and again on April 22. - March 31 announcement eased collateral requirements to borrow from the CBRT and opened the door for unlimited bond purchases where it was stated that “. . . limits might be revised depending on market conditions.” - CBRT and BRSA introduced several financial repression measures increasing banking system risk exposure, encouraging banks to lend at low rates or buy government bonds. - Capital flow management measures reduced domestic banks’ reserve requirements for foreign currency deposits and put limits on the daily amounts of domestic banks’ swap transactions. - Note: effective and transparent communication of policy actions is critical for market perception. - Fiscal policy and stimulus size: - Initial government package announced on March 18 was 2 percent of GDP; scope expanded as conditions evolved. - Minister of Finance and Treasury announced on May 29 that pandemic related government expenditure has already reached 260 billion TL. - Even with revised numbers, the package still remains around 5 percent of GDP. - Comparative context: - Average size of fiscal stimulus among G20 countries: about 10 percent of GDP. - Germany: 32 percent of GDP. - Turkey: small, lagging behind 16 of the G20 countries. - Direct transfers and targeting: - Transfer payments channels include: transfer payments for needy families; enhanced employment protection; temporary ban on layoffs with state subsidy; unemployment insurance benefits. - These transfer payments add up to 12 billion TL, which is only about 4.6 percent of the total stimulus package. - Policy trade-off: targeted transfers vs broader “helicopter money” in presence of sizable informal sector; feasibility depends on funding availability. ### Key scenarios, model findings, and parameter values (epidemiology & macro outcomes) - Epidemiological model structure (SIR with sectoral extension): - ∆S_t = −β S_{t−1} I_{t−1} / N - ∆R_t = γ I_{t−1} - ∆I_t = β S_{t−1} I_{t−1} / N − γ I_{t−1} - ∆S_t + ∆R_t + ∆I_t = 0 - Industry-specific infection rates: β_i = β_0 Prox_i for i = 1, … , K. - Lockdown scenario outcomes (model results): - No lockdown baseline: - Pandemic peak around the 150th day with a total toll of around 14 million infections. - Virus taken under control after approximately 300 days. - Under no lockdown: 1.13 percent of the population dies (assuming a 1.5 percent mortality rate). - GDP declines 11.0 percent. - Partial lockdown (industries operate as usual; βi unchanged), 240 days starting day 10: - Reproduction number declines below 2 but remains above 1; containment is not achieved. - If β0 = 0.5×β0: GDP declines 11.6 percent. - If β0 = 0.25×β0: GDP declines 10.9 percent. - If β0 = 0.1×β0: GDP declines 10.5 percent. - After lockdown removal (day 250), low-infection-rate scenarios see rapid increases leading to peaks within 50 days. - Concluding implication: partial lockdown may need to continue indefinitely without an efficient drug or vaccine. - Extended partial lockdown (full year): - Flattens the curve; additional economic costs hover around 0.5 percent of GDP compared to shorter partial lockdown. - Full lockdown: - Full lockdown contains virus fastest and lasts approximately 40 days in benchmark. - Full lockdown (R0 = 0) example: contained within 39 days; GDP decline about 5.8 percent. - Less effective full lockdown (R0 = 0.02): duration 54 days; GDP declines 7.6 percent. - Costs of delaying full lockdown: - One-day delay: infections increase by more than 10,000; lockdown extended by two more days; GDP cost increases to 5.9 percent. - Two-day delay: lockdown duration increases to 43 days; GDP decline 6.2 percent. - One-week delay: GDP decline 7.3 percent. - Comparative observation: costs of full lockdown are lower than any partial lockdown scenarios considered. - Mortality comparisons across scenarios: - Effective full lockdown: 0.001 percent of the population dies. - No lockdown: 1 percent of the population dies. - Partial lockdowns lasting 250 days: about 0.8 percent of the population dies. - Partial lockdown extended to a full year: deaths decline to about 0.5 percent of the population. ### Sectoral breakdown of economic costs and role of external finance - Role of trade linkages: - Sectors with higher share of imports in intermediate inputs more exposed to supply-chain disruptions: paper products; computers and electronics; electrical equipment; rubber and plastic; coke and refined petroleum; pharmaceuticals. - Sectors with higher share of exports in total output more exposed to adverse foreign demand: motor vehicles; transportation equipment; electrical equipment; computer and electronics; accommodation and food services (tourism-related). - Scenario-specific top sectoral VA losses (selected entries): - No lockdown (Scenario 1, β=0.14): top VA losses among 15 industry groups: - Accommodation and food services: 36% - Arts, entertainment, recreation and other service activities: 33% - Real estate activities: 20% - Full lockdown (Scenario 2, 91-131, R0=0): top VA losses: - Accommodation and food services: 12% - Construction: 9.5% - Mining and non-quarrying of non-energy producing products: 9.1% - Partial lockdown (Scenario 3, 10-250, 0.25×β0): top VA losses: - Accommodation and food services: 36% - Arts, entertainment, recreation and other service activities: 34% - Real estate activities: 21% - Supply vs demand drivers: - No lockdown: demand channel drives output on almost all days until containment. - Full lockdown: supply channel drives output during closures; demand dominates ~30 days before restrictions. - Partial lockdown: supply dominates first 100 days; demand drives output thereafter and during post-removal peaks. - Regression evidence on external finance and sectoral losses (summary of Table 1): - Regression: ∆Yi = β0 + β1 I-O Tradei + β2 I-O Trade Financei + εi; robustness includes FX (foreign currency debt ratio). - Coefficients (heteroskedastic-consistent SEs in parentheses): - I-O Trade: 15.98** (6.402); 16.49** (6.426); 16.03** (6.388); 16.53** (6.412) - I-O Trade Finance: 34.78* (17.351); 35.63** (17.256); 34.95* (17.331); 35.80** (17.234) - FX: 0.16** (0.076) (included in Columns (2) and (4)) - Observations: 33; R2: 0.12 and 0.2 across specifications. - Interpretation: stronger I-O trade linkages and sectors that finance I-O trade via capital flows exhibit larger output/VA losses; higher sectoral FX debt associated with larger losses. - Policy implication: - EMs with high external debt and domestic FX debt, such as Turkey, risk a prolonged sudden-stop-type recession if risk appetite towards EMs does not return soon. - Reforms to enhance productivity, transparency, institutional strength, and market mechanisms could help attract global liquidity. ### Taking stock: timing, second waves, and policy trade-offs - Observed country paths: - Full lockdown examples: Greece, New Zealand, Denmark — policy minimizing economic costs by effective containment. - No lockdown: very few countries; associated with much higher mortality despite possible lower short-run economic costs. - No lockdown followed by full lockdown (e.g., UK): delayed full lockdown yields higher mortality and economic cost than an earlier full lockdown. - Partial lockdowns evolving into full lockdowns (e.g., Italy, France, Germany, Spain): later full lockdowns lasted longer than necessary. - Enhanced partial lockdown (Turkey): phased measures, age-based curfews, weekend restrictions in 31 largest cities; after ~45 days R0 reduced below 1 and new patients < recovered patients by last week of April. - Key findings on timing and demand: - Early full lockdown can control pandemic relatively quickly; country performance depends on recovery/infection rate determinants. - Demand modeled as function of infections and credit card spending: demand will not normalize merely by removing restrictions if infections remain sizable. - Forward-looking demand (expectations-based) could change outcomes; leaders can affect expectations via policy. - Second-wave and border reopening trade-offs: - Early full-lockdown countries face second-wave risk if borders open; closed borders keep them in extended partial lockdown with amplified costs for open economies. - If a second wave hits, immediate and potentially global lockdown would be most effective. - Fiscal timing and production capacity risks: - Lockdowns increase short-term costs but can yield faster recovery and long-term gains. - Model assumes productive capacity remains intact; absence of comprehensive support risks liquidity turning into solvency problems, bankruptcies, deeper recession, and sluggish recovery. - Estimated stimulus needs earlier presented should be interpreted as lower-bound costs to keep economic units alive. - Quick implementation of stimulus that compensates income loss can minimize long-term damage to production capacity. ### QE, debt monetization, credibility, and recommended transparency measures - Technical distinctions and risks: - Monetary financing (helicopter money) = central bank prints money and transfers to firms/households or purchases government bonds directly. - QE = central bank buys sizeable government bonds; difference from debt monetization hinges on ability to drain liquidity later. - In advanced economies with well-anchored inflation, QE vs monetization distinction easier to ascertain; in EMs with weak credibility distinction blurs. - Badly managed QE can erode credibility, de-anchor inflation expectations, and cause sharp currency depreciation. - Application to Turkey: - Rapid increase in CBRT’s bond holdings; purchases limited to 10% of CBRT’s balance sheet but March 31 statement allowed limits revision. - No explicit QE program with a clear exit strategy announced. - Market concern: central bank purchases approaching “5 or 10 per cent of GDP” could imply potential for considerable currency depreciation. - Recommended transparency measures: - Communicate a detailed bond purchase calendar with spending targets and conditions for draining money. - Consider a Special Purpose Vehicle (SPV) to buy government bonds through it, separating COVID-19 related purchases from daily monetary policy. - Ensure money generated is spent in targeted sectors and announced by the government. - Policy conclusion: - Even well-executed QE likely insufficient for countries like Turkey; joint planning of fiscal needs and capital flows required given foreign financing needs. ### External funding needs, capital controls, and role of external anchors - External funding and FX exposure: - Total amount of external debt that needs to be paid or rolled-over in 2020 is 23 percent of GDP. - Current open FX position of the entire corporate sector as of January 2020: -$175 billion (almost 25 percent of GDP). - Turkey’s upcoming external debt payment in 2020: $169 bn (23 percent of GDP). - Observed capital flows and liquidity lines: - Large EMs saw sizable capital outflows (example Brazil): $12 billion outflow from stock market and $19 billion from bond market before May 2020. - Average emergency swap agreement granted by the IMF: about $11bn. - Total outstanding amount granted by the Federal Reserve’s international swap lines: $18.9 bn as of April 14. - On May 20, Turkey announced expansion of existing swap line with Qatar to $15 billion. - Policy alternatives and tradeoffs: - Tightening monetary policy (rate hikes) may backfire during risk-off shocks, especially with low policy credibility. - Loosening monetary policy necessary given negative demand shock but EMs with high external debt cannot rely on rate cuts entirely. - FX liquidity via swap agreements can help but may be insufficient if pandemic weakens firms’ debt servicing. - Debt moratorium on foreign private creditors would involve complex restructuring and could be disorderly if not synchronized. - Capital controls: benefits and risks - Controls to trap foreign currency assets may have unintended consequences; historically effectiveness limited. - EMs often need more inflows, not outflow controls; breaching contracts can provoke legal action and harm future access. - Capital controls can erode credibility and deter needed foreign capital; removal can be protracted (example: Iceland took 10 years post-2007-2008). - Recent domestic measures perceived as mild capital controls (limiting TL supply in swap markets; notifying govt on sizable FX transfers; restricting TL transactions of custodian banks; May 24 tax increase from 0.2 percent to 1 percent on exchange rate and gold transactions) risk deterring foreign investors. - Historical precedent and role of IMF: - 2001 crisis response: securities transfers and CBRT asset expansion (~8 percent of GDP in securities purchases); coordinated reforms and 2002 IMF standby supported confidence and disinflation. - Scenario-based estimated costs and policy implications (Turkey calibration): - Model yields estimated annual cost of COVID-19 crisis: between 5.8 percent and 11 percent of Turkish GDP depending on effectiveness and duration of lockdown. - Most cost-effective full lockdown scenario implies a quarterly GDP contraction of 17 percent. - Policy implications: - Transparent and clearly communicated monetary policy critical if QE-type programs finance domestic debt. - Well-designed macroeconomic and structural reform package increases credibility and lowers external borrowing costs. - Funding from international institutions signals support and helps cover financing gaps. - Global coordination in containing pandemic and synchronizing lockdowns would speed recovery and minimize economic costs. *Source: wpiea2020133-print-pdf - Section 8 concludes.* ### Section 8 concludes. ### wpiea2020133-print-pdf - Section 8 concludes. ### COVID-19 literature and the paper's contribution - Rapidly growing literature integrates infection dynamics into macroeconomic analysis, often via SIR models or extensions. - Representative findings from cited studies: - Stock (2020) and Alvarez et al. (2020): optimal policy is a full lockdown covering the majority of the population with restrictions removed gradually afterwards. - Acemoglu et al. (2020): multi-risk SIR model by age groups; targeted measures such as full lockdown for the elderly could be more effective. - Alon et al. (2020): developing-country perspective with market distortions and an informal sector; conclude full economy lockdowns are infeasible and lockdowns targeting the elderly might be better. - Farboodi et al. (2020): individuals sharply reduce activity in laissez-faire; socially optimal response imposes severe early restrictions. - Eichenbaum et al. (2020): incorporating supply and demand with government-altered activity via a consumption tax; find severe early containment is socially optimal. - Krueger et al. (2020): introduce differential transmission rates based on consumption or employment choice; highlight importance of sectoral heterogeneity. - Barrot et al. (2020), Bonadio et al. (2020), Baqaee et al. (2020): incorporate input-output (I-O) linkages to analyze propagation of sectoral supply shocks. - Baqaee and Farhi (2020b): Keynesian model with sectoral supply shocks and aggregate demand shocks; show large amplification in a network economy (absent infection dynamics). - Guerrieri et al. (2020): multi-sector economy where supply shocks can turn into larger aggregate demand shocks (no infection dynamics). - Baqaee and Farhi (2020a): in a non-SIR general equilibrium with nonlinear production networks, nonlinearities amplify the impact of COVID-19 between 20 to 100 percent; reductions in labor supply are less costly if spread across many sectors. - Distinctive contribution: - First paper (to the authors' knowledge) to incorporate both supply and demand shocks at the sector level for an open economy operating within the international production network. - Estimates COVID losses considering domestic and foreign sectoral shocks and their amplification through I-O linkages. - Links economic losses to I-O trade links and capital flows empirically to inform fiscal needs of emerging markets (EMs). ### Theoretical framework and lockdown scenarios - Model schematic elements: - SIR model with industrial heterogeneity (teleworkable vs on-site workers; physical proximity). - Supply channel: proportion of workforce impaired by infection, dependent on teleworkable share and physical proximity. - Demand channel: consumption changes linked to infection levels and mitigated by lockdowns. - Inter-industry and international linkages determine propagation and equilibrium output (minimum of supply- and demand-implied output). - Lockdown scenarios implemented: - partial lockdown: - all industries remain open; - teleworkable portion of employees work from home; - low infection rate for teleworkables and the general public; - high infection rate for on-site workers. - full lockdown: - only essential industries remain open; - workers in non-essential sectors stay at home; - infection rates lowered for almost everyone. - Data and identification choices: - Use Dingel and Neiman (2020)’s list of teleworkable occupations to capture teleworkable employment share by industry. - For on-site professions, assume viral transmission depends on physical proximity between workers or between workers and customers. - Estimate proportion of workforce impaired during the pandemic using teleworkable share and physical proximity measures within the SIR model. - Demand modeling: - Two demand scenarios: normal times and brunt of the pandemic. - Proxy for peak-pandemic demand shocks: credit card purchase data from Central Bank of the Republic of Turkey (CBRT). - For sectors missing credit card data, use other real-time demand proxies. - Pre-COVID demand: consumption spending from Turkish national accounts. - Reduced-form function models adjustment of demand from pre-COVID levels to the lowest COVID shock level as a function of number of infected people; demand normalizes as infections decline. ### Open-economy implications and global coordination - Global synchronization of lockdowns: - If lockdowns are globally synchronized, pandemic is controlled faster. - Faster global decline in infections leads to quicker normalization of domestic and foreign demand and minimizes economic costs. - Equilibrium: - Final industry-level production is the minimum of supply-implied and demand-implied outputs. ### Initial conditions in Turkey and external financing needs - Pre-pandemic macro background and vulnerabilities: - Since 2017, inflation rose while Turkish Lira (TL) depreciated. - August 2018 marked an exchange rate crisis triggered by political tension with US; rapid TL depreciation pushed FX-indebted firms toward bankruptcy. - Growth rate in Q1 2020: 4.5 percent. - Unemployment rate declined to 12.7 percent. - As of first week of April 2020, net reserves of CBRT: $26 billion, of which $25 billion was borrowed from domestic banks. - IMF-defined budget deficit excluding one-time transfers: close to 5 percent of GDP (average over last 5 years). - Current account deficit: around 2.5 percent of GDP (average over last 5 years). - External debt and currency composition: - Turkey’s external debt: 60 percent of GDP at end-2019. - Historical context: - 2001 total external debt: 57 percent of GDP (public sector 24 percent, private sector 22 percent). - By 2019, total external debt comparable to 2001 with 56 percent of GDP; private sector share 36 percent of GDP, public debt 21 percent of GDP. - As of December 2019, almost 60 percent of total external debt is denominated in USD. - Maturity and short-term rollover needs: - Total external debt: $437 billion. - Short-term portion: $124 billion (17 percent of GDP). - Of short-term debt, $93 billion held by private sector. - BIS data: $96 billion of total external debt belongs to the banking system. - External debt needing rollover in 2020: $169 billion (approximately 23 percent of GDP). - Banking sector’s share in short-term debt: $81 billion. - Rollover scenarios: - If rollover ratios stay at current levels, Turkey needs around $30 billion in 2020. - If rollover ratios fall to GFC levels, Turkey might need around $90 billion in 2020. - Note: out of total external debt, $437 billion and the short-term $124 billion and $169 billion rollover figures are preserved verbatim from source. - Market sentiment and capital flows: - From beginning of year until week of April 24, 2020: - $2.7 billion of equity held by foreign investors sold to domestic investors. - $5.5 billion of government bonds held by foreign investors sold to domestic investors. - Corporate bond sell-off from January 3 to April 24, 2020: $86.5 million outflow (negligible relative to bank loans and government bonds). - Cross-border bank loan rollovers and banking sector flows: - In January 2020, Turkish banks paid net $0.8 billion more in short-term loans than they borrowed, and $0.7 billion more in long-term loans. - In April 2020, Turkish banks paid $1 billion in long-term loans. - Turkish banks had been net payers in external long-term loans over recent periods. - FX exposure amplifies COVID-19 economic costs: - More than 1/3 of total external funding is obtained through bank loans in Turkey, almost all in FX. - Half of entire corporate sector debt is in FX and most is borrowed from domestic banks. - Sectors with higher FX exposure experience larger output losses under the COVID-19 shock (correlation reported: corr: 0.558 & t-stat: 2.231). - Figure 5 (descriptive): plots 2016 sectoral FX exposure (foreign currency debt / total debt) against estimated sectoral output loss (%) under a no-lockdown, no-policy scenario; shows stronger FX funding → sharper output declines. - Policy implications highlighted by initial conditions: - High FX corporate exposure and sizable external debt imply that exchange rate depreciation acts as an amplification mechanism exacerbating bankruptcies and output losses in FX-exposed sectors. - Well-designed policies needed to address loan restructuring, non-performing loans, and non-viable firms given private-sector FX debt concentration. *Source: wpiea2020133-print-pdf - Section 8 concludes.* ### 3.2 Policy Response to COVID-19 ### 3.2 Policy Response to COVID-19 ### Monetary and financial policies - CBRT cut rates by 100 basis points immediately during their emergency meeting on March 18, 2020 and again on April 22. - March 31 announcement eased collateral requirements to borrow from the CBRT and opened the door for unlimited bond purchases where it was stated that “. . . limits might be revised depending on market conditions.” - CBRT and BRSA introduced several financial repression measures in the following days that increase the risk exposure of the banking system, encouraging banks to lend at low rates or buy government bonds. - Capital flow management measures were introduced that reduced domestic banks’ reserve requirements for foreign currency deposits and put limits on the daily amounts of domestic banks’ swap transactions. - Note on communications: for an EM, market perception of such measures is critical because potential risks and external borrowing costs are priced by global investors. Effective and transparent communication of policy actions is as critical as the actions themselves. ### Fiscal policy and stimulus size - The stimulus package announced by government on March 18 is consistent with general frameworks adopted by other countries (postponement of tax obligations, social security premiums and credit payments for companies in the services sector; increased limits of the Credit Guarantee Fund; temporary income support for workers whose companies ceased production; cash assistance for needy families). - The original package announced on March 18 was announced to be 2 percent of GDP; the scope was expanded as conditions evolved. - The Minister of Finance and Treasury announced on May 29 that the pandemic related government expenditure has already reached 260 billion TL. - Even with revised numbers, the package still remains around 5 percent of GDP. - Comparative context: - Average size of fiscal stimulus among G20 countries is about 10 percent of GDP. - Germany: 32 percent of GDP. - Turkey: small, lagging behind 16 of the G20 countries, reflecting limited fiscal space of EMs relative to advanced economies. ### Direct transfers and targeting - Transfer payments channels in the stimulus package include: transfer payments for needy families; enhanced employment protection by loosening short-term work allowance rules; a temporary ban on layoffs with a state subsidy for affected workers; unemployment insurance benefits. - These transfer payments add up to 12 billion TL, which is only about 4.6 percent of the total stimulus package. - Policy trade-off highlighted: Should transfer payments cover only those who lost their income or be “helicopter money” in a context with a sizeable informal economy? The answer depends on funding availability. A generous program that does not trigger longer term macroeconomic imbalances can minimize economic damage and prevent long-term risks; a less comprehensive program would delay recovery. ### Key scenarios, model findings, and parameter values - Lockdown scenarios and outcomes: - Full lockdown contains the virus fastest and lasts for approximately 40 days. - Partial lockdown cannot contain the virus within a year. - Economic costs of a partial lockdown are significantly higher than full lockdown because duration increases substantially. - Mortality outcomes: - Full lockdown: only 0.002 percent of the population dies in a well implemented full lockdown. - Partial lockdown: mortality ranges between 0.32 to 0.96 percent. - The analysis does not quantify economic costs of lost lives; incorporation of those costs would further strengthen the superiority of full lockdown. - Epidemiological model structure (SIR and sectoral extension): - Conventional SIR equations: - ∆S_t = −β S_{t−1} I_{t−1} / N - ∆R_t = γ I_{t−1} - ∆I_t = β S_{t−1} I_{t−1} / N − γ I_{t−1} - ∆S_t + ∆R_t + ∆I_t = 0 (population constant) - Sectoral heterogeneity: - Economy composed of K sectors indexed by i = 1, … , K with L_i workers and non-working population N_NW. - Each industry i has TW_i teleworkable workers and N_i on-site workers, with L_i = TW_i + N_i. - At-home group (i = 0) population: N_0 = N_NW + sum_{i=1}^K TW_i. - Industry-specific infection rates: β_i = β_0 Prox_i for i = 1, … , K, where Prox_i is the proximity index for industry i. - Susceptible at-home infection: ∆S_{0,t} = −β_0 S_{0,t−1} I_{t−1} / N where I_t = sum_{i=1}^K I_{i,t} + I_{0,t}. - On-site susceptible dynamics: ∆S_{i,t} = −β_i S_{i,t−1} I_{i,t−1} / N_i − β_0 S_{i,t−1} I_{t−1} / N. - Recovery common across groups: ∆R_{i,t} = γ I_{i,t−1}. - Change in infected: ∆I_{i,t} = −(∆R_{i,t} + ∆S_{i,t}). - Calibration and parameter values: - Recovery rate γ = 1/14 ≈ 0.07 (mean recovery time of 14 days). - World Health Organization reported R_0 between 2 and 2.5; the lower end, R_0 = 2, is used. - In absence of industrial heterogeneity, R_0 = 2 and γ = 0.07 implies β = 0.14. - With industrial heterogeneity, employment-weighted average β_i’s are matched to β. For an on-site worker in industry i, implied β parameter approximated by (β_0 + β_i); for a non-working individual the parameter is β_0. - Imposed condition: - β_0 N_0 / N + sum_{i=1}^K (β_0 + β_i) N_i / N = β - Leading to solved expression: β_0 = β ( 1 + sum_{i=1}^K Prox_i N_i / N )^{−1} with β = 0.14. - Production specification: - Output linear in labor: Y_{i,t} = Z_i L_{i,t}. - During pandemic, available workers: - ˜L_{i,t} = (N_{i,t} − I_{i,t}) + TW_i (1 − I_{0,t} / N_0) - Pandemic-period sector output: - Y^S_{i,t} = Z_i ˜L_{i,t}. *Source: wpiea2020133-print-pdf - 3.2 Policy Response to COVID-19* ### 4.3 Demand ### 4.3 Demand ### Demand shifts during the pandemic - Consumer behavior changes due to: - Fear of infection, linked to the number of infected individuals. - Fear of transmitting the disease to others, with high risk from asymptomatic cases. - Uncertainty about the duration of the pandemic and the economic outlook, which reduces aggregate expenditure. - Direct income effects from layoffs or sharp declines in demand for output. - Two demand profiles are considered: one for normal times (determined from consumption data in national accounts) and one for the COVID-19 period (estimated from credit card spending data; industry reports and expert opinions used where credit card data are unavailable). - Individuals move smoothly between the two profiles as a function of the number of infected individuals; final good consumption is mapped back to industry output via the input-output framework. ### Modeling demand: utility specification and demand parameters - Representative agent utility: Cobb-Douglas form U(e1, ..., en) = ∏_{i=1}^n e_i^{α_i}, with ∑_{i=1}^n α_i = 1 and 0 < α_i < 1. - Under normal times, expenditure in industry i is e_i = α_i w. - During the pandemic, preferences adjust and the utility becomes ̃U(e1, ..., en, I) = ∏_{i=1}^n e_i^{̃α_i(I)}, with: - ̃α_i(I) = α_i for I ≤ 0.1 ̄I. - lim_{I→∞} ̃α_i(I) ≡ ̄α_i with ∑_{i=1}^n ̄α_i = 1 and 0 < ̄α_i < 1. - Income channel: available income for expenditure decreases by a ratio of 1−η(I) compared to normal times. - η(I) is a decreasing function of I and satisfies η(I) = 1 for I ≤ 0.1 ̄I. - lim_{I→∞} η(I) = ̄η with 0 < ̄η ≤ 1. - Minimum income necessary for survival at the brunt of the pandemic is ̄η × w (achievable via transfers). - Calibration detail: - Scaling parameter ̄I is set to 50,000 to capture a relevant range for the number of infections in the Turkish context. - The limit I ≤ 0.1 ̄I therefore corresponds to I ≤ 5,000. ### Demand ratio δ_i(I): combining preference and income channels - Expenditure in industry i during infections I relative to normal times is defined by: - δ_i(I) = ̃α_i(I) η(I) / α_i. - Limiting cases: - For small I (I ≤ 0.1 ̄I), δ_i(I) = 1. - For large I (I → ∞), ̄δ_i = ̄α_i ̄η / α_i. - If demand completely collapses for industry i, ̄δ_i = 0 (e.g., airlines). - If no change in demand, ̄δ_i = 1 (e.g., food). - ̄δ_i represents the utmost demand change for sector i under a fully developing pandemic (globally valid limit). - Functional form for smooth transition: - δ_i(I) = 1 if I ≤ 0.1 ̄I. - δ_i(I) = ̄δ_i / [1 + (I/̄I − 0.1) ̄δ_i + (I/̄I − 0.1)] if I > 0.1 ̄I. - The inverse hyperbolic functional form yields a smooth transition between limiting cases; marginal impact of infections changes at a rate inversely proportional to I. - Alternative explicit specifications for η(I) and ̃α_i(I) (for i = 1, ..., n): - η(I) = 1 and ̃α_i(I) = α_i if I ≤ 0.1 ̄I. - η(I) = ̄η / [1 + (I/̄I − 0.1) ̄η + (I/̄I − 0.1)] and ̃α_i(I) = (̄α_i / α_i) ̄η + (I/̄I − 0.1) ̄δ_i + (I/̄I − 0.1) (for I > 0.1 ̄I) as presented in the Appendix. ### Mapping demand changes to final demand and output - Final demand in country c, industry i: F_{c,i}. - Revised final demand during the pandemic when infections are I_t: - ̃F_{c,i}(I_t) = F_{c,i} δ_i(I_t). - To map final demand changes to industry output while accounting for international linkages, the OECD Inter-Country Input-Output (ICIO) Tables (2015) are used. - Dataset structure: 36 industries × 69 entities (65 countries) → matrix of 2484 × 2484 entries. - Direct requirements matrix A obtained by dividing ICIO rows by total output of industry j; A summarizes inputs needed per $1 of output. - Input-output equilibrium relations: - Y = A Y + F leads to Y = (I − A)^{-1} F. - Total output of country c: Y_c = ∑_{i=1}^n Y_{c,i}. - Demand-driven output during infections I_t: Y^D_t = (I − A)^{-1} ̃F(I_t). ### Equilibrium allocation and GDP computation - Equilibrium output during the pandemic: element-wise minimum of supply-driven output and demand-driven output: - Y^{EQ}_t = min(Y^S_t, Y^D_t). - Value-added in industry i, country c at time t: - VA^{EQ}_{t,c,i} = Y^{EQ}_{t,c,i} (VA_{c,i} / Y_{c,i}). - Country GDP at time t: - GDP^{EQ}_{t,c} = ∑_{i=1}^n VA^{EQ}_{t,c,i}. ### Data sources and measurement details - OECD ICIO Tables for 2015 used; industrial classification: OECD aggregation of 2-digit ISIC Rev 4 codes to 36 sectors (labeled as OECD ISIC Codes). - Teleworkable share and physical proximity measures: - Occupational composition used; teleworkable occupations from Dingel and Neiman (2020). - Physical proximity from O*NET Work Context "Physical Proximity" values. - O*NET categories divided into 3; values > 1 if reported category ≥ 3 (Slightly close) and a single occupation proximity value created by weighting normalized scores by response shares. - Industry-level values obtained by weighting occupation values with Occupational Employment Statistics (OES) from U.S. BLS (four-digit NAICS); converted to OECD ISIC via the correspondence table between 2017 NAICS and ISIC Revision 4. - Teleworkable share and proximity index provided in Table A.2 of the Appendix. - Employment data from Turkish Social Security (SGK) Agency (four-digit NACE Revision 2); converted to 36 OECD ISIC codes using Eurostat correspondence table; public administration employment supplemented with President’s office data where SGK lacks coverage. - Demand change estimation: - Publicly available credit card spending data from the CBRT used to estimate sectoral demand changes; matching of CBRT spending data to OECD ISIC industries is provided in Table A.5 of the Appendix. - For sectors without credit card data, projections based on sector reports, other countries’ experiences, and historical sector/manufacturing data used. - Aggregate demand shock magnitude: 23% when focusing only on sectors with credit card data; 16% when considering the full set of sectors. - Sensitivity analysis indicates little or no qualitative change in findings. - Under full lockdown: - Active industries identified using the Turkish Ministry of Interior decree of April 10, 2020; supplemented with food sector and household and sanitary goods. - Shares of OECD ISIC industry activity during lockdown computed from 4-digit employment data (list of sectors in Table A.4 of the Appendix). - Share of public employees not affected by lockdown calculated using publicly available information (Table A.6 of the Appendix). *Source: IMF Working Paper (section 4.3 "Demand"), wpiea2020133-print-pdf* ### 5.1 Infection Rates under Alternative Lockdown Scenarios ### 5.1 Infection Rates under Alternative Lockdown Scenarios ### Framework and assumptions - Changes are imposed on β0 (infection rate of the non-working population) and possibly on βi (infection rate of working population in industry i) to simulate pandemic trajectories. - Decline in β reflects lockdown effectiveness and depends on country characteristics (demographics, culture, influence of scientific committees, media) and recovery rate (quality of healthcare services and ICU capacity). - The pandemic is considered successfully contained if the number of total infections declines to 5000 after observing the peak. - For containment operations, the model assumes that for each infected individual, we need to test ten additional people on average (thus 5000 patients imply about 50,000 tests). - Reference benchmark reproduction rate: R0 = 2 in the no-lockdown scenario. ### No lockdown baseline outcomes - Pandemic peak around the 150th day with a total toll of around 14 million infections. - Virus taken under control after approximately 300 days. - Under no lockdown: - 1.13 percent of the population dies (assuming a 1.5 percent mortality rate). - GDP declines 11.0 percent. ### Partial lockdown (industries operate as usual; βi unchanged) - Partial lockdown reduces β0 only. Three reduction cases considered relative to reference setting: 0.5×β0, 0.25×β0, 0.1×β0. - Lockdown implemented for 240 days (starting early on the 10th day, active until the 250th day) in Figure 8. - Under the three 240-day partial lockdown scenarios: - Reproduction number declines below 2 but remains above 1 in all three scenarios; containment is not achieved. - If β0 = 0.5×β0 (high infection rate; red line): GDP declines 11.6 percent. - If β0 = 0.25×β0 (moderate; green line): GDP declines 10.9 percent. - If β0 = 0.1×β0 (low; black line): GDP declines 10.5 percent. - Epidemiological dynamics after lockdown removal (day 250): - All three scenarios have approximately the same number of infections at removal. - Low infection rate scenarios (green and black) see rapid increases in new cases, leading to peak levels within 50 days after lockdown removal. - High infection rate and no-lockdown scenarios (blue and red) show a steady decline after removal because more people were infected during lockdown and thus acquired immunity. - Concluding implication: In absence of an efficient drug or vaccination, a partial lockdown may need to continue indefinitely until cases decline to 5000. ### Extended partial lockdown (full year) - Figure 9 simulates partial lockdown lasting a full year with industries operating as usual (βi unchanged). - Main advantage: flattens the curve by spreading infections over time and allowing for a larger recovery rate. - Economic costs: - Additional economic costs of the longer partial lockdown hover around 0.5 percent of the GDP compared to the shorter partial lockdown. - The limited added costs arise because demand decline reaches a maximum early and successive production reductions mainly reflect supply decline due to increased infections. ### Full lockdown scenarios - Figure 10: full lockdown when infections ≈ 80,000. - Fully effective full lockdown (R0 = 0; blue line): - Pandemic contained within 39 days (gray shaded area). - GDP decline about 5.8 percent. - Less effective full lockdown (R0 = 0.02; yellow shaded area): - Lockdown duration increases to 54 days. - GDP declines 7.6 percent. - Costs of delaying full lockdown (Figure 11; benchmark = full lockdown starting 91st day): - One-day delay: - Number of infections increases by more than 10,000. - Lockdown needs to be extended by two more days (red line). - GDP cost increases to 5.9 percent. - Two-day delay (green line): - Lockdown duration increases to 43 days. - GDP decline is 6.2 percent. - One-week delay (black line): GDP decline is 7.3 percent. - After 100 days, virus starts to spread again and prematurely ending the lockdown is ineffective. - Comparative observation: - Costs of full lockdown are lower than any of the partial lockdown scenarios considered. ### Mortality comparisons across scenarios - Effective full lockdown: 0.001 percent of the population dies. - No lockdown: 1 percent of the population dies. - Partial lockdowns lasting 250 days: about 0.8 percent of the population dies. - Partial lockdown extended to a full year: deaths decline to about 0.5 percent of the population. *Source: 5.1 Infection Rates under Alternative Lockdown Scenarios (wpiea2020133-print-pdf)* ### 5.3 Sectoral Breakdown of Economic Costs ### 5.3 Sectoral Breakdown of Economic Costs ### The role of trade linkages in sectoral costs - International linkages affect sectoral costs through trade relationships and capital inflows. - Figure 13 observations (OECD ICIO Tables): - Sectors with higher share of imports in intermediate inputs (left panel) likely more exposed to supply-chain disruptions: paper products; computers and electronics; electrical equipment; rubber and plastic; coke and refined petroleum; pharmaceuticals. - Sectors with higher share of exports in total output (right panel) are more exposed to adverse foreign demand: motor vehicles; transportation equipment; electrical equipment; computer and electronics; tourism-related services such as accommodation and food services. - Supply shocks are not explicitly modeled but are implicitly captured through changes in final demand. ### Sectoral breakdown of economic costs under alternative lockdown scenarios - Three benchmark scenarios (as defined in figures and notes): - Scenario 1: No lockdown, β=0.14. - Scenario 2: Full lockdown, 91-131, R0=0. - Scenario 3: Partial lockdown, 10-250, 0.25×β0. - General findings across scenarios: - Teleworkable or essential sectors are less severely affected (examples: education, IT, public administration). - Non-essential or on-site work sectors are more severely affected (examples: accommodation and food services; arts, entertainment and recreation; construction). - Scenario-specific sectoral losses (value-added loss, VA loss (%)): - No lockdown (Scenario 1): highest losses among 15 industry groups: - “Accommodation and food services”: 36% - “Arts, entertainment, recreation and other service activities”: 33% - “Real estate activities”: 20% - Full lockdown (Scenario 2): highest losses among 15 industry groups: - “Accommodation and food services”: 12% - “Construction”: 9.5% - “Mining and non-quarrying of non-energy producing products”: 9.1% - Partial lockdown (Scenario 3): highest losses among 15 industry groups: - “Accommodation and food services”: 36% - “Arts, entertainment, recreation and other service activities”: 34% - “Real estate activities”: 21% - Interpretation of supply vs demand drivers (Figure 15 summary): - No lockdown (Scenario 1): demand channel drives output on almost all days until containment; supply channel prevails only in the early days (not shown). - Full lockdown (Scenario 2): supply channel drives output during closure of non-essential industries; demand channel prevails approximately 30 days before restrictions are implemented. - Partial lockdown (Scenario 3): supply channel dominates in the first 100 days; demand drives output for the rest of the year, including days with new peaks after premature removal of partial lockdown. - Sectoral heterogeneity in dominance of channels: - Under no lockdown and partial lockdown scenarios, demand channel prevails longer in non-essential sectors (e.g., accommodation & food services; arts & entertainment; real estate). - Under full lockdown, supply channel dominates for almost all sectors after restrictions, except “Human health & social work” and “Public administration.” ### The role of external finance in sectoral cost - Sector-level regression specification: - I-O Trade Financei = sum_{c=1}^{n} ((((Exports_{c,i} − Imports_{c,i})/Output_{i}) × Capital Flows_{c})/n) - Regression: ∆Yi = β0 + β1 I-O Tradei + β2 I-O Trade Financei + εi (Equation (37)) - ∆Yi: economic cost of the COVID-19 shock for sector i (percentage change in overall economic activity proxied by output or value added during pandemic relative to pre-pandemic level). - I-O Tradei: I-O trade linkage for sector i. - I-O Trade Financei: sector-level proxy capturing interdependence between trade linkages and external finance needs (weighted sum of net I-O trade of country-sector pairs where weights are country-specific capital flows divided by number of countries). - Robustness check: add FX (ratio of foreign currency debt in total debt as of 2016) to capture domestic foreign-currency borrowing by sector. - A priori expectations: - Tradable sectors expected to be hit harder due to exposure to foreign demand shocks. - Even non-tradable sectors can be hit via I-O linkages. - Sectors relying more on external borrowing expected to face larger economic costs due to increased risk aversion. - Regression results (Table 1 summary; heteroskedastic-consistent standard errors in parentheses; ***, **, * indicate significance at 1%, 5%, 10% respectively): - Dependent variable: Output Loss (Columns 1-2) and VA Loss (Columns 3-4) - Coefficients: - I-O Trade: 15.98** (6.402); 16.49** (6.426); 16.03** (6.388); 16.53** (6.412) - I-O Trade Finance: 34.78* (17.351); 35.63** (17.256); 34.95* (17.331); 35.80** (17.234) - FX: 0.16** (0.076); 0.16** (0.076) (appears in Columns (2) and (4) as additional explanatory variable) - Number of observations: 33 in all columns - R2: 0.12; 0.2; 0.12; 0.2 (corresponding to Columns (1)-(4)) - Interpretation of regression findings: - Positive and statistically significant coefficients confirm importance of international linkages. - Sectors with stronger I-O links suffer larger COVID-19 related losses. - Sectors that finance strong production links through capital flows and sectors with higher FX exposure suffer even more. - Implication: EMs with high external debt and domestic FX debt, such as Turkey, face risk of a prolonged sudden-stop-type recession if risk appetite towards EMs does not return soon. - Policy implication: a well-designed plan involving reforms to enhance productivity, transparency, institutional strength, and free market mechanisms could help attract abundant global liquidity injected by advanced economies during the COVID-19 crisis. *Source: wpiea2020133-print-pdf - 5.3 Sectoral Breakdown of Economic Costs* ### 5.4 Taking Stock ### 5.4 Taking Stock ### Paths adopted by countries during the pandemic - Full lockdown examples: Greece, New Zealand and Denmark — identified as the policy that minimizes economic costs by containing the pandemic most effectively. - No lockdown: Very few countries; may yield lower economic costs but results in a significantly higher death toll; economic costs mostly depend on changes in demand. - No lockdown followed by a full lockdown: Example — UK. Analysis indicates that a non-delayed full lockdown would have led to less mortality and lower economic costs because it would begin with a smaller number of infections. - Partial lockdown followed by full lockdown: Examples include Italy, France, Germany, Spain, Iran, Russia. Several announced gradual lifting of restrictions. Full lockdown duration is longer than it could have been if implemented earlier. Example: Italy — full lockdown went into effect on March 10 and restrictions announced to be removed by May 4 after approximately two months under full lockdown. - Enhanced partial lockdown: Turkey started immediate partial measures, enhanced over time: - Schools closed on March 16; businesses encouraged to work remotely where possible. - Curfew for people above the age of 65 and those with chronic diseases on March 21. - Curfew extended to those younger than 20 on April 5, effectively putting close to 40% of the population under full lockdown. - Full lockdown on weekends and national holidays starting April 9 in 31 largest cities which constitute approximately 87% of the population. - After about 45 days since beginning enhanced partial lockdown measures, R0 is reduced below 1 and the number of new patients is lower than the number of recovered patients as of the last week of April. - Note: the 31 cities include the 30 metropolitan municipalities and Zonguldak, which constitute close to 79% of the population; with age-based restrictions in the rest of Turkey this increases the number close to 87%. ### Key findings on timing, lockdowns, and demand - A full lockdown at early stages can bring the pandemic under control relatively quickly; some countries implemented this successfully while others (e.g., India) did not succeed despite early full lockdown. - Country performance depends on factors affecting recovery and infection rates. - Two-month evaluation of Turkey’s lockdown measures indicates Turkey did reasonably well; potential contributing factors: - remarkable ICU capacity - young population - less care homes - generally compliant population where government decrees are not challenged - If an enhanced partial lockdown lowers R0 below 1, full lockdown may not be imminent; however, results indicate lockdown duration would have been shorter if more restrictive measures were adopted immediately. - Demand is modeled as a function of the number of infections and combined with actual spending decline measured with credit card purchases. Implication: demand will not normalize merely by removing restrictions if the number of infections remains sizable; reopening risks infections rising again. - The model assumes actual infections highly correlated with demand; a forward-looking demand curve (function of infection expectations) could allow leaders to affect expectations and revive demand by removing restrictions — reducing the negative demand effect. ### Second-wave, borders, and reopening trade-offs - Countries that implemented early full lockdowns and controlled the pandemic face a second-wave risk if they open borders. - If borders remain closed, these countries cannot fully normalize and may suffer an extended partial lockdown with amplified economic costs for open economies. - Takeaway: if a second wave hits, an immediate and potentially global lockdown would be most effective. ### Fiscal policy, stimulus timing, and production capacity risks - Lockdowns increase short-term costs but can increase long-term gains via faster recovery. - Model shortcoming: does not incorporate damage to productive capacity from company closures — assumes productive capacity remains intact and companies resume production post-pandemic (an optimistic assumption). - Without comprehensive support, liquidity issues may become solvency issues leading to bankruptcies, deeper recession, and sluggish recovery. - Estimated stimulus needs presented earlier should be interpreted as lower bound costs of a stimulus package necessary to offset COVID-19 damages and keep economic units alive. - Quick implementation of stimulus that compensates income loss and enables faster recovery would minimize long-term damage to production capacity. - Delayed stimulus leads to more company failures, more layoffs, further demand decline, more bankruptcies, and elevated economic costs that become unmanageable. - Fiscal transfers can help ensure supply chains are not destroyed and economic units remain functional and ready to resume production once the pandemic is contained. - In Emerging Markets (EMs), fiscal space is limited; policy options are constrained — “We do not live in whatever it takes region, we can do whatever we can.” (Mauricio Cardenas) ### Quantitative Easing (QE) versus debt monetization: distinctions and risks - Monetary financing (a.k.a. "helicopter money") = central bank prints money and transfers resources to firms and households directly or indirectly (e.g., purchasing government bonds). - QE: central bank prints money and buys sizeable amounts of government bonds; balance sheet enlarges via direct loans or large-scale asset purchases. Advantage of direct lending: liquidity is drained more easily when loans are repaid. - Technical difference between open market money printing and debt monetization is slim; QoE policies can be seen as debt monetization in the short run. Federal Reserve distinguishes debt monetization as “permanent” funding for government by the central bank and separates QE from debt monetization. - Criterion used by the Federal Reserve: bond purchases are debt monetization if the central bank fails to drain the money later and money remains permanently, causing inflationary pressures. - In advanced economies with well-anchored inflation (example: inflation rate has not exceeded the 2 percent target in the US or Europe after large-scale QE), distinction is easier to ascertain. - In EMs, distinction gets blurrier for countries with a history of high inflation and weak credibility. - Key to successful QE: policy credibility. Badly managed QE erodes credibility, de-anchors inflation expectations, pushes inflation higher, and causes sharp currency depreciations. If not executed properly and money is not drained at the right time, QE can turn into inflationary debt monetization. - EM central banks with weaker track records must clearly communicate QE policies for credibility. ### Application to Turkey and transparency considerations - Rapid increase in CBRT’s bond holdings reflects sizable balance sheet expansion. Size of purchases currently limited to 10% of CBRT’s balance sheet; statement on March 31 suggests these limits might be adjusted as needed. - No explicit QE program with a clear exit strategy has been announced as typically done in advanced economies. - Transparent communication about limits and data-driven ties to economic conditions is important to influence foreign investors’ sentiments. - Market commentary cited: if central bank purchases approach “5 or 10 per cent of GDP [for countries like South Africa, Indonesia or Turkey], then you are getting into the realm of the danger zone . . . there would be the potential for considerable currency depreciation.” - Countries with credible, independent central banks (e.g., Chile or those in eastern Europe per cited view) are better placed to deal with pressures because investors fear less that central banks will enable excessive government spending. - If inflation is low (example cited: 1.5 percent inflation rate as in the US) and a deep recession is imminent, inflationary consequences of QE may not be immediate because public expects the central bank to drain money later and maintain price control. Market participants’ belief that the government will not default keeps interest rates under control. - Turkey has missed its 5 percent inflation target for some time and market sentiment indicates eroded policy credibility. - Goal: convince market participants that QE will not turn into inflationary debt monetization. Recommended transparency measures: - Communicate a detailed bond purchase calendar with spending targets and conditions for draining money from the system. - Consider a Special Purpose Vehicle (SPV) to allow central banks to buy government bonds through the SPV, separating COVID-19 related bond purchases from daily monetary policy maintenance. This would make monetary expansion due solely to COVID-19 easily trackable. - Ensure money generated through such programs is spent in targeted sectors and announced by the government. - Even well-executed QE is likely insufficient for countries like Turkey; joint planning of fiscal needs and capital flows is required given foreign financing needs. *Source: wpiea2020133-print-pdf — 5.4 Taking Stock.* ### 6.2 External Funding Needs, Capital Controls, and the Role of External Anchor ### 6.2 External Funding Needs, Capital Controls, and the Role of External Anchor ### External funding needs and sectoral vulnerabilities - Turkey faces large external funding needs: - Total amount of external debt that needs to be paid or rolled-over in 2020 is 23 percent of GDP. - The current open FX position of the entire corporate sector as of January 2020 is -$175 billion (almost 25 percent of GDP). - Turkey’s upcoming external debt payment in 2020 is as large as 23 percent of the GDP ($169 bn). - Sectors with stronger trade linkages and higher external funding needs are more vulnerable during the COVID-19 crisis. - Rapid increase in risk premia for most EMs raises the cost of external borrowing and makes rollover harder. ### Observed capital flows and liquidity lines - Large EMs have already seen sizable capital outflows (example: Brazil): - 12 billion USD outflow from the stock market before May 2020. - 19 billion USD outflow from the bond market before May 2020. - Swap and emergency liquidity metrics cited: - The average emergency swap agreement granted by the IMF is about $11bn. - Total outstanding amount granted by the Federal Reserve’s international swap lines is $18.9 bn as of April 14. - On May 20, Turkey announced that the existing swap line with Qatar is expanded to $15 billion. - The Federal Reserve did not expand the list of countries eligible for a swap line since the GFC; Turkey was not in the original list during GFC. ### Policy alternatives and tradeoffs - Tightening monetary policy (rate hikes) to compensate for risk premium: - May backfire during risk-off shocks, especially in countries with low policy credibility and high risk premia. - Tight monetary policy may not be fully effective under a large risk-off shock. - Loosening monetary policy to provide accommodation: - Necessary given large negative demand shock, but EMs with high external debt cannot rely on rate cuts entirely. - EMs must balance supporting domestic demand and limiting domestic currency volatility. - FX liquidity via swap agreements with the Federal Reserve or other international institutions: - Can help address liquidity needs arising from COVID-19 but may be insufficient if the pandemic extends and weakens firms’ ability to service debt. - Debt moratorium on foreign private creditors: - Would involve complicated debt default and debt restructuring. - Unless synchronized offers from private creditors occur, a disorderly moratorium would hamper medium- to long-term credibility. ### Capital controls: potential benefits and risks - Proposal: introduce capital controls to trap foreign currency assets in Turkey, limiting TL depreciation. - Historical and practical considerations: - Capital controls on outflows during large risk-off shocks might have unintended consequences and historically have not been very effective. - EMs most likely need more capital inflows, not controls on outflows. - Breaching contracts may provoke legal action by private creditors and compromise future market access. - Panic by foreign investors invested in local currency bonds could put additional pressure on the local currency and inflation. - Capital controls can erode policy credibility and scare foreign capital needed during recovery; once in place, controls on outflows can take a long time to remove (example: Iceland took 10 years to lift controls after 2007-2008 crisis). - Potential limited benefit: - Capital controls could be argued to prevent further dollarization triggered by TL liquidity injected through QE, but dollarization can also be prevented if policy credibility keeps inflation expectations anchored. ### Recent domestic measures perceived as mild capital controls - Measures mostly on domestic residents, examples: - Limiting TL supply in international swap markets. - Notifying the government regarding sizable FX transfers abroad. - Restricting TL transactions of large custodian banks. - On May 24, tax on exchange rate and gold transactions increased from 0.2 percent to 1 percent. - Market perception risks: - Such measures deter foreign investors by limiting capital mobility and creating impressions of random legislative changes. - Unpredictable regulatory interventions can discourage future capital inflows and damage policy credibility. ### Historical precedent: 2001 episode and IMF involvement - 2001 crisis response elements: - State banks and Savings Deposit Insurance Fund (SDIF) faced significant losses; government securities were transferred to these institutions and then sold to CBRT to receive cash. - Size of securities purchases reached approximately 8 percent of GDP during that time. - CBRT’s total assets in real terms increased about 122 percent from January 2000 to November 2001. - CBRT drained excess liquidity gradually through conventional methods to prevent unintended declines in market rates. - 2002 standby agreement with the IMF: - Met external funding needs and provided credibility to boost confidence and prevent excessive depreciation. - Comprehensive package of reforms helped limit domestic funding needs and the volume of asset purchases, reducing pressure on inflation expectations. - Public finance and debt management laws improved fiscal transparency and accountability; additional funds for banking system restructuring were offset by increasing public savings elsewhere to keep the overall budget under control. - Coordinated efforts supported a successful disinflation performance without exchange rate volatility. ### Sectoral importance of FX credit flow - Sectors that rely more on external borrowing are hit harder during COVID-19; keeping flow of FX credit is particularly important for maintaining production capacities. ### Scenario-based economic costs and policy implications (from model estimates) - Model calibration to Turkey yields estimated annual cost of COVID-19 crisis: - Between 5.8 percent and 11 percent of Turkish GDP depending on effectiveness and duration of lockdown. - Most cost-effective full lockdown scenario: - Implies a quarterly GDP contraction of 17 percent. - Policy implications: - Transparent and clearly communicated monetary policy is critical if QE-type programs are used to finance domestic debt. - A well-designed and comprehensive package of macroeconomic and structural reforms will increase policy credibility and reduce external borrowing costs and spreads. - Funding from international institutions would signal support for the policy package and help cover financing gaps, lowering external finance premia and easing financial strains. - Global coordination in containing the pandemic and synchronizing lockdowns would speed recovery and minimize economic costs; absence of coordination raises the role of all policy options for EMs. *Source: IMF working paper section 6.2 and related sections as provided.* ### References ### References (wpiea2020133-print-pdf) ### Key citations and topics covered - Network origins of aggregate fluctuations and production networks: Acemoglu et al., “The Network Origins of Aggregate Fluctuations,” Econometrica, 2012, 80(5), 1977–2016. - Multi-risk SIR and targeted lockdowns: Acemoglu, Chernozhukov, Werning, and Whinston, Working Paper 27102, National Bureau of Economic Research May 2020. - Stochastic epidemic models and numerical simulation: Allen, Linda JS, Infectious Disease Modelling, 2017, 2(2), 128–142. - Policy responses for developing countries to COVID-19: Alon et al., Working Paper 27273, National Bureau of Economic Research 2020. - Simple planning problem for COVID-19 lockdown: Alvarez, Argente, and Lippi, Working Paper 26981, National Bureau of Economic Research 2020. - Consumer responses using transaction data: Andersen et al., CEPR Discussion Papers 14809, C.E.P.R. Discussion Papers 2020. - Fed monetization of government debt: Andolfatto and Li, 2013. Economic Synopses, Federal Reserve Bank of St. Louis. - Deadly debt crises in emerging markets during COVID-19: Arellano, Bai, and Mihalache, Working Paper 27275, National Bureau of Economic Research 2020. - Estimating COVID-19 fatality rate difficulties: Atkeson, Working Paper 26965, National Bureau of Economic Research 2020. - Gross capital flows by sector: Avdjiev, Hardy, Kalemli-Ozcan, and Serváçn, Working Paper 23116, National Bureau of Economic Research January 2017. - COVID-induced economic uncertainty: Baker et al., Working Paper 26983, National Bureau of Economic Research 2020. - Economics in the time of COVID-19: Baldwin and Weder di Mauro, Vol. 26, Centre for Economic Policy Research, London, 2020. - Nonlinear production networks and COVID-19: Baqaee and Farhi, Working Paper 27281, National Bureau of Economic Research 2020; related work on disaggregated Keynesian economies Working Paper 27152. - Reopening scenarios: Baqaee, Mina, and Stock, Working Paper 27244, National Bureau of Economic Research 2020. - Sectoral effects of social distancing: Barrot, Grassi, and Sauvagnat, 2020. Available at SSRN. - Critical community size for measles and epidemic thresholds: Bartlett, M. S., Journal of the Royal Statistical Society: Series A (General), 1960, 123(1), 37–44. - Capital-account liberalization signaling: Bartolini and Drazen, The American Economic Review, 1997, 87(1), 138–154. - Integrated monetary and financial policies for small open economies: Basu et al., in NBER, Summer Institute 2020 International Finance & Macroeconomics, 2020. - Is coronavirus as deadly as reported? Bendavid and Bhattacharya, Wall Street Journal, 2020, 24. - Global supply chains in the pandemic: Bonadio et al., Working Paper 27224, National Bureau of Economic Research 2020. - Capital flows and risk-taking channel: Bruno and Shin, Journal of Monetary Economics, 2015, 71, 119–132. - Country- and institution-specific technical reports and working papers on pandemic response, testing, mobility, production networks, social distancing, monetary policy spillovers, debt overhang, bankruptcy, and data gaps: multiple entries including CEPR, NBER, arXiv preprints, VOX CEPR Policy Portal, Science, Journal of Clinical Medicine, IMF Policy Discussion Papers, and other working papers and mimeos cited in the reference list. ### Epidemiology, measurement, and data sources - Substantial undocumented infection and rapid dissemination: Li et al., Science, 2020. - Incubation period and epidemiological characteristics with right truncation: Linton et al., Journal of Clinical Medicine, 2020, 9(2), 538. - New antibody blood tests and measurement of pandemic scale: Vogel, Science, March 19 2020. - Data gaps and policy response: Stock, Working Paper 26902, National Bureau of Economic Research 2020. - Antibody testing and identifying immune workers as a priority for restarting the economy: Dewatripont et al., VOX CEPR Policy Portal 2020. --- ### Appendix (A) ### List of Figures and Tables (as provided) - Figure A.1: The Structure of OECD Inter-Country Input-Output Table - Figure A.2: Sovereign Bond Issuance in Turkey - Table A.1: Fiscal Responses to the COVID-19 Shock in the G20 Countries - Table A.2: Proximity Index and Teleworkable Share Across Industries - Table A.3: Demand Changes Across Industries - Table A.4: List of the Lockdown Sectors - Table A.5: CBRT Credit Card Spending Titles Corresponding to OECD ISIC Sectors - Table A.6: List of the Active Sectors in Public Administration during full lockdown ### Figure A.1: Structure of OECD Inter-Country Input-Output Table (ICIO) - Represents breakdown of output corresponding to 36 industries and 69 countries, giving a matrix of 2484×2484 entries. - In any industry-country combination: - Output (Y) equals intermediate use (Z) plus final demand (F). - Final demand components across 69 countries: - fd1: Households Final Consumption Expenditure (HFCE) - fd2: Non-Profit Institutions Serving Households (NPISH) - fd3: General Government Final Consumption (GGFC) - fd4: Gross Fixed Capital Formation (GFCF) - fd5: Change in Inventories and Valuables (INVNT) - fd6: Direct purchases by non-residents (NONRES) - fd7: Statistical Discrepancy (DISC) - Value added + taxes - subsidies on intermediate products (VA) included; Output (Y) and Output (Z)(F)(Y) referenced in table notes. - Industry list referenced in Table A.2. ### Table A.1: Fiscal responses to the COVID-19 shock in the G20 countries (excerpted entries) - Presentation format: Country | % GDP | Explanation - Argentina - 3 - Adopted measures (totaling about 3.0 percent of GDP, 1.2 percent in the budget and 1.8 percent off-budget, based on authorities’ estimates) - Australia - 10.8 - Total expenditure and revenue measures of A$194 billion (9.9 percent of GDP). The Commonwealth government has committed to spend almost an extra A$5 billion (0.3 percent of GDP). State and Territory governments also announced fiscal stimulus packages, together amounting to A$11.5 billion (0.6 percent of GDP) - Brazil - 6.5 - The authorities announced a series of fiscal measures adding up to 6.5 percent of GDP. Public banks are expanding credit lines for businesses and households, with a focus on supporting working capital (credit lines add up to over 3 percent of GDP), and the government will back a 0.5 percent of GDP credit line to cover payroll costs. - Canada - 8.4 - Key tax and spending measures (8.4 percent of GDP, $193 billion CAD). - China - 3.8 - An estimated RMB 2.6 trillion (or 2.5 percent of GDP) of fiscal measures or financing plans have been announced. The overall fiscal expansion is expected to be significantly higher, reflecting the effect of already announced additional measures such as an increase in the ceiling for special local government bonds of 1.3 percent of GDP. - France - 19 - The authorities have announced an increase in the fiscal envelope devoted to addressing the crisis to e110 billion (nearly 5 percent of GDP, including liquidity measures), from an initial e45 billion included in an amending budget law introduced in March. A new draft amending budget law has been introduced on April 16. This adds to an existing package of bank loan guarantees and credit reinsurance schemes of e315 billion (close to 14 percent of GDP). - Germany - 31.6 - The federal government adopted a supplementary budget of e156 billion (4.9 percent of GDP). The government is expanding the volume and access to public loan guarantees for firms of different sizes and credit insurers increasing the total volume by at least e757 billion (23 percent of GDP). In addition to the federal government’s fiscal package, many state governments (Länder) have announced own measures to support their economies, amounting to e48 billion in direct support and e73bn in state-level loan guarantees (Authors: Another 3.7% of GDP). - India - 1.1 - Finance Minister Sitharaman on March 26 announced a stimulus package valued at approximately 0.8 percent of GDP. These measures are in addition to a previous commitment by Prime Minister Modi that an additional 150 billion rupees (about 0.1 percent of GDP). Numerous state governments have also announced measures thus far amount to approximately 0.2 percent of India’s GDP. - Indonesia - 2.8 - In addition to the first two fiscal packages amounting to IDR 33.2 trillion (0.2 percent of GDP), the government announced a major stimulus package of IDR 405 trillion (2.6 percent of GDP) on March 31, 2020. - Italy - 26.4 - On March 17, the government adopted a e25 billion (1.4 percent of GDP) ‘Cura Italia’ emergency package. On April 6, the Liquidity Decree allowed for additional state guarantees of up to e400 billion (25 percent of GDP). - Japan - 21.1 - On April 7 (partly revised on April 20), the Government of Japan adopted the Emergency Economic Package Against COVID-19 of Y *Source: wpiea2020133-print-pdf - References* ### 117.1 trillion (21.1 percent of GDP) ### wpiea2020133-print-pdf - 117.1 trillion (21.1 percent of GDP) ### COVID-19 fiscal packages by selected countries (key figures and measures) - Mexico - Up to 180 billion pesos (0.7 percent of 2019 GDP) requested from Congress. - Austerity program to free up 2.5 percent of GDP for additional health expenditures and priority investment. - Listed headline: 117.1 trillion (21.1 percent of GDP) appears as a document marker. - Republic of Korea - Direct measures amount to 0.8 percent of GDP (approximately KRW 16 trillion). - March 24: financial stabilization plan of KRW 100 trillion (5.3 percent of GDP). - April 22: additional KRW 35 trillion (1.8 percent of GDP). - April 22: key industry stabilization fund KRW 40 trillion (2.1 percent of GDP). - Russian Federation - Total cost of the fiscal package estimated at 2.1 percent of GDP. - Saudi Arabia - SAR 70 billion ($18.7 billion or 2.8 percent of GDP) private sector support package (announced March 20). - Reduce spending in non-priority areas of the 2020 budget by SAR 50 billion (2.0 percent of GDP). - Use of unemployment insurance fund (SANED) for wage benefits: SAR 9 billion (0.4 percent of GDP). - Temporary electricity subsidies (SAR 0.9 billion) and increased health sector resource support SAR 47 billion. - South Africa - Listed figure: 0.2 (percent of GDP). Reference: https://www.globalpolicywatch.com/2020/04/south-africas-economic-response-to-the-covid-19-pandemic/ - Spain - Key measures about 1.6 percent of GDP, e18 billion (amount could be higher depending on usage and duration). - Extended up to e100 billion government guarantees for firms and self-employed. - Additional funding for ICO credit lines: e10 billion. - Special ICO credit line for tourism: e400 million. - Turkey - TL100 billion package announced: TL75 billion ($11.6 billion or 1.5 percent of GDP) in fiscal measures; TL 25 billion ($3.8 billion or 0.5 percent of GDP) doubling credit guarantee fund. - Gradually increased to be 5% of GDP. - United Kingdom - Policy measures adding £86 billion in 2020-21. - Coronavirus business interruption loan scheme announced as up to £330 billion of support for businesses. - Source cited: https://obr.uk/coronavirus-reference-scenario/ - United States of America - US$484 billion Paycheck Protection Program and Health Care Enhancement Act. - Estimated US$2.3 trillion (around 11% of GDP) CARES Act. - US$8.3 billion Coronavirus Preparedness and Response Supplemental Appropriations Act. - US$192 billion Families First Coronavirus Response Act. - These together provide around 1% of GDP (in addition to the larger packages). - Notes: - Table reports COVID-19 relief packages (as percent of GDP) by the G20 and details of fiscal packages. - Source: IMF Policy Tracker unless otherwise noted. Access Date: April 29, 2020. ### Sovereign bond issuance in Turkey (quarterly external bonds issued vs due) - Graphic description (2000q1 to 2020q2): - Vertical axis: in billion USD. - Series shown: Bonds issued, Bonds due, Crisis spell, Crisis x ratio<1. - Crisis spells highlighted in grey. - Red areas where Turkey experienced a financial crisis and rollover ratio fell below 1: 2001q2, 2008q4, 2018q2, 2020q2. - Heading: Rollover risk, bonds issued, bonds due — Turkey. ### Industry-level proximity index and teleworkable share (OECD ISIC) - Table entries (OECD ISIC Definition — Proximity — Teleworkable Share). Selected entries: - 01T03 Agriculture, forestry and fishing — 0.86 — 0.06 - 05T06 Mining and extraction of energy producing products — 1.08 — 0.32 - 10T12 Food products, beverages and tobacco — 1.12 — 0.13 - 26 Computer, electronic and optical products — 1.03 — 0.54 - 62T63 IT and other information services — 1.01 — 0.88 - 64T66 Financial and insurance activities — 1.02 — 0.79 - 85 Education — 1.22 — 0.86 - 86T88 Human health and social work — 1.28 — 0.35 - 58T60 Publishing, audiovisual and broadcasting activities — 1.11 — 0.69 - Notes on construction: - Proximity index derived from O*NET occupation-level physical proximity categories normalized with category (3) as benchmark (divided by 3). - Teleworkable share uses Dingel and Neiman (2020) list of teleworkable occupations. - Industry-level values are weighted averages of occupation values using BLS Occupational Employment Statistics (OES) and converted from NAICS to ISIC Revision 4. ### Demand changes across industries (Turkey sectoral demand change percentages and explanations) - Format: OECD ISIC Definition — Change — Explanation (selected rows) - 01T03 Agriculture, forestry and fishing — 100% — Based on projections on the manufacturing sector using data on credit card spending from the database of CBRT and computations using historical data. - 13T15 Textiles, wearing apparel, leather and related products — 50% — Based on estimates using data on credit card spending from the database of CBRT. - 16 Wood and products of wood and cork — 90% — Based on projections on the manufacturing sector using CBRT credit card spending and historical computations. - 22 Rubber and plastic products — 90% — Based on projections on the manufacturing sector using CBRT credit card spending and computations using historical data. - 29 Motor vehicles, trailers and semi-trailers — 70% — Based on other countries’ experiences and sectoral reports. - 35T39 Electricity, gas, water supply, sewerage, waste and remediation services — 100% — No change. - 45T47 Wholesale and retail trade; repair of motor vehicles — 110% — Based on estimates using CBRT credit card spending. - 55T56 Accommodation and food services — 25% — Based on estimates using CBRT credit card spending. - 61 Telecommunications — 100% — Based on estimates using CBRT credit card spending. - 62T63 IT and other information services — 100% — Based on other countries’ experiences and sectoral reports. - 64T66 Financial and insurance activities — 100% — Based on estimates using CBRT credit card spending. - 84 Public admin. and defence; compulsory social security — 125% — Median Package size 5%. Public spending is close to %20 of GDP. - 86T88 Human health and social work — 100% — Based on other countries’ experiences and sectoral reports. - Notes: - Demand changes estimated using publicly available data and CBRT credit card spending; categorized by OECD ISIC Codes. ### Lockdown sectors (Turkey, NACE Rev. 2 lists) - Panel A: Lockdown Sectors (selected entries) - 01 Crop and animal production, hunting and related service activities - 10 Manufacture of food products - 35 Electricity, gas, steam and air conditioning supply - 36 Water collection, treatment and supply - 46 Wholesale of pharmaceutical goods - 86 Human health activities - 87 Residential care activities - Panel B: Additional Sectors (selected entries) - 10 Manufacture of food products - 46 Wholesale of food, beverages and tobacco - 47 Retail sale in non-specialised stores with food, beverages or tobacco predominating - Notes: - List based on the decree issued by the Turkish Ministry of Interior on April 10, 2020. The lockdown was effective for only two days and covers those in Panel A; Panel B supplements the list. ### CBRT credit card spending titles mapped to OECD ISIC sectors - Selected concordances (CBRT Definition — OECD ISIC Code) - Car Rental — 69T82 - Car Rental-Sales/Service/Parts — 45T47 - Petrol Stations — 19 - Various Food — 10T12 - Education/Stationary — 45T47 - Electric & Electronic Goods, Computers — 26 - Accomodation — 55T56 - Telecommunication — 61 - E-commerce Transactions — 62T63 - Government/Tax Payments — 84 - Notes: - Concordance used to match CBRT credit card spending titles with OECD ISIC Codes. ### Active subsectors in Public Administration during full lockdown (Turkey) - Type and counts (selected entries) - Public (All) — 2820095 — Source: http://www.sbb.gov.tr/kamu-istihdami/ - Security — 273000 — Source: https://tr.wikipedia.org/wiki/EmniyetGenelM%C3%BCd%C3%BCrl%C3%BC%C4%9F%C3%BC - Gendarmerie — 150000 — Source: https://www.jandarma.gov.tr/jandarma-genel-komutanligi-2019-yili-faaliyet-raporu - Health — 642184 — Source: https://www.saglik.gov.tr/TR,11588/istatistik-yilliklari.html - Share: 37.77% (share of the active sub-sectors in the entire sector) - Notes: - Table provides list of occupations in Public Administration that work during full lockdown with counts and sources. *Source: wpiea2020133-print-pdf (IMF PDF content provided).* --- _Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020133-print-pdf.pdf_