## tfotppwea

## Source details

**Canonical URL:** [tfotppwea](https://www.imf.org/-/media/files/publications/dp/2021/english/tfotppwea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2021/english/tfotppwea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2021/english/tfotppwea.pdf.json)

---

### Executive Summary
- COVID-19 caused a sudden stop in travel and a collapse in economic activity worldwide.
- Tourism accounts for more than 10 percent of the global economy and is a major exporter of foreign exchange for many countries.
- The pandemic put millions of tourism jobs at risk, with informal and migrant workers—particularly women and youth—suffering disproportionately.
- Many tourism dependent economies entered the pandemic with limited fiscal space, inadequate external buffers, and foreign exchange revenues extremely concentrated in tourism.
- Strong correlation found between the spread of COVID-19 and big data high-frequency indicators on travel, suggesting healthcare system quality will be pivotal in recovery.
- Recovery opportunity: accelerate shift toward eco-sustainable tourism (lower density, higher value-added, greater digitalization).

### The role of tourism before COVID-19
- Tourism became one of the world’s most important growth engines: more than 10 percent of global GDP directly and indirectly.
- International tourists rose to more than 1.5 billion by end-2019 from 680 million in 2000.
- Spending by international tourists amounted to nearly US$1.6 trillion as of end-2019.
- Domestic tourism comprised more than 70 percent of total tourism spending in 2019.
- Domestic tourism spending rose from US$2.2 trillion in 2005 to US$4.5 trillion by 2019; China accounts for over a fifth of such spending.
- On average, tourism directly accounts for about 3½ percent of global GDP; total contribution (direct, indirect, induced) is estimated to exceed 10 percent of global GDP.
- Focus on tourism-dependent countries (TDCs): contribution of tourism and related sectors above 10 percent of GDP during 2016–18.

### Employment and structural characteristics
- Tourism employed above 300 million globally.
- Tourism generates 1 out of 10 youth jobs (WTTC, 2019).
- Women account on average for 54 percent of those employed in the tourism sector compared to 39 percent in the broader economy.
- Sector characteristics: low-skill intensity, high seasonality, high turnover, high informality; actual employment may be significantly larger than official figures.
- Migrant-worker shares cited: 16 percent of tourism workers in the European Union and 20 percent in the United States are foreigners; in Thailand, a fifth of hospitality workers come from neighboring countries.
- ILO ASEAN survey: nearly 97 percent of unemployed migrants in the tourism sector have no access to their host country’s social safety nets (ILO 2020).

### Immediate COVID-19 impacts and mobility indicators
- International flight numbers declined by 46 percent year over year by October 2020.
- Collapse in international tourist arrivals (UNWTO preliminary data, as of August 2020):
  - Fall of 70 percent in the first eight months of 2020; 80 percent drop in August vs same period in 2019.
  - UNWTO (2020) suggests a drop in tourist arrivals of 78 percent in 2020 would translate into a loss in visitor spending of US$1.2 trillion and place more than 100 million direct tourism jobs at risk.
- Example: EU tourist nights by domestic travelers fell about 40 percent vs 70 percent by nonresidents as of July 2020 relative to previous year.
- Community mobility average decline relative to pre-pandemic baseline: as much as 47 percent in Asia-Pacific and 75 percent in the Western Hemisphere.
- Google Trends: travel-related searches collapsed after a March spike, partially recovered in summer, then depressed in the second wave.

### Empirical findings (gravity-model results)
- Augmented gravity model (38,184 country pairs, 1995–2017) finds infectious diseases impact tourism flows more in developing countries than advanced economies.
- Disease-specific effects (2SLS-IV selected coefficients):
  - Ebola: Origin = -0.065*** [0.005]; Destination = -0.089*** [0.008]
  - SARS: Origin = -0.387*** [0.092]; Destination = -0.078*** [0.104]
  - Malaria and yellow fever: smaller/insignificant effects in many specifications.
- Example interpretation: for SARS, a 10 percent increase in confirmed cases leads on average to a reduction of 4.7 percent in international tourist arrivals.
- Scaling SARS coefficients to COVID-19 prevalence yields an approximate decline of 82.5 percent in international tourism flows globally (noted as an upper-bound estimate); actual impact in first half 2020 was -65 percent.

### GIMF modeling and scenario design
- Model: IMF’s Global Integrated Monetary and Fiscal model (GIMF) with a tourism module; annual, multi-region, micro-founded general equilibrium model.
- Tourism productivity scarring assumption: productivity in tourism sector assumed 10 percent lower than before for a given allocation of capital and labor (benchmark); historical evidence suggests 6 to 15 percent post-epidemic productivity declines.
- Benchmark scenario:
  - Near-total shutdown of tourism in Year 1; gradual recovery with permanent −10 percent productivity shock (AA_S).
  - International tourism resumes in second year; returns near pre-pandemic level after three years under equitable vaccine distribution expectations.
- Downside scenario:
  - Sectoral productivity reduction of −20 percent; slower recovery of preferences (decay rates increased); tourists differentiate by healthcare quality via NTBs.
- Upside scenario:
  - No productivity scarring (0 percent); faster preference recovery (decay rates reduced), temporary surge in 2021 if vaccine rollout permits.
- Regional impacts (benchmark GIMF simulation):
  - Pacific Islands and Caribbean hardest hit (real GDP falls by more than 20 percent in tourism-dependent regions).
  - Net importers of tourism services (Latin America and China) may see increases in real GDP despite tourism shocks.
- Key model technical points preserved:
  - Real exchange rate adjusts immediately in first year ("jumper").
  - Fiscal policy: governments follow pre-COVID-19 spending plans; automatic stabilizers active.
  - Monetary policy: passive, inflation-targeting rules; no additional monetary stimulus to offset tourism shocks in baseline.

### Policy framework: three phases and considerations
- Phase 1, crisis mitigation:
  - Fiscal support to buttress demand and preserve jobs; scope for well designed fiscal stimulus mindful of fiscal space and debt sustainability.
  - Monitor and support stressed firms early to prevent liquidity stress turning into insolvency; banks to facilitate restructuring for viable firms, possibly with central bank support.
  - Caution against relaxing loan classification or provisioning rules.
- Phase 2, reopening:
  - Emphasize health and hygiene protocols; incentivize domestic tourism; consider COVID-free travel bubbles where feasible.
  - Targeted policies for youth and women, digitalization to broaden inclusion and mitigate scarring.
- Phase 3, recovery:
  - Shift to eco-sustainable, low-density, higher value-added tourism to reduce health risks and diversify economies.
  - Transition from liquidity support to corporate restructuring; calibrate policies to country heterogeneity; provide job training and support for retooling.

### Fiscal stimulus scenarios (ASEAN-5 application; illustrative for Pacific Islands relevance)
- Two fiscal stimulus packages over three years: 3 percent of GDP for two years and 1.5 percent of GDP in third year.
- Lower-multiplier package focuses on public consumption, transfers to all households, cuts in labor income taxes.
  - First-year real GDP loss under this package: 1.6 percent.
  - Cumulative deficit: 7.5 percent of GDP.
  - Debt-to-GDP ratio increases by 6.6 percent of GDP.
- Higher-multiplier package focuses on infrastructure investment, transfers targeted to poorer, liquidity-constrained households, cut in VAT.
  - First-year real GDP loss under this package: 0.8 percent.
  - Debt-to-GDP ratio increases by 6.1 percent of GDP.
- Many Pacific Island states have limited fiscal space; development-partner financing and tourism-specific fiscal support may be necessary.

### Travel-bubble simulations (ANZ–PIC example)
- Two-simulation approach:
  1. Counterfactual worse: tourism shut down for first two years (instead of one).
  2. Travel bubble: ANZ and PIC reopen tourism internally and to each other in Year 2 before general reopening in Year 3.
- Bubble technical parameters (Year 2):
  - Remove ALPHA_C shock in ANZ and PIC; return ALPHA_SH with decay rates 0.80 in ANZ and 0.65 in PIC.
  - For Year 2 only, add NTBs on imports of tourism services by ANZ from other regions of 200 percent; similarly for PIC.
- Outcomes:
  - Pacific Islands gain very large GDP gains due to tourism’s outsize role.
  - Australia and New Zealand see modest GDP gains mostly from reviving investment.
  - Bubble reduces risk of COVID-19 cases from outside tourists (benefit not quantified in model).

### Box 2 recommendations: travel bubble implementation, health strengthening, digitalization, and diversification
- Strengthen TDC health systems: early detection, treatments, capacity building, accreditation of facilities to attract health tourism; global vaccine production/distribution as a public good emphasized.
- Promote low-density and eco-sustainable tourism: multi-sectoral strategy embedded in national development frameworks; Costa Rica and Thailand cited as examples.
- Accelerate technological innovation and digitalization: contact tracing, touchless service delivery, digital concierge; reskilling critical to manage automation impacts.
- Strengthen linkages and diversify exports: lower average tariffs "about 10 percent"; example: St. Kitts and Nevis per capita tourism exports ~ US$7,500 in 2019 (about three times higher than Malaysia’s manufacturing exports per capita); ECCU micro-states could offset more than a 10 percent permanent decline in tourism exports if manufacturing per capita reached that level.
- Upgrade quality across sectors and pursue export complexity where feasible; constraints exist for countries lacking complex export baskets (Cambodia, Dominican Republic, Honduras, Jamaica, Nicaragua, Mongolia, Panama).

### Pacific Islands: modeled impacts, fiscal dynamics, and policy sequencing
- Real GDP falls by more than 20 percent in tourism-dependent regions (Pacific Islands and Caribbean).
- Larger tourism share produces larger wealth shock and stronger REER depreciation in Pacific Islands; stronger net export response (lower imports/higher exports).
- Fiscal dynamics:
  - Many Pacific governments accumulated significant debt pre-pandemic; benchmark assumes deterioration in fiscal positions and sharp increase in debt due to growth and exchange-rate shocks.
  - Assumes government discretionary spending unchanged; automatic stabilizers small.
  - If affordable financing replaced by cuts in government consumption or social transfers: fiscal position would not improve; private consumption would be substantially weaker; negative impacts on real GDP would increase the debt-to-GDP ratio.
- Policy sequencing and measures:
  - Immediate: coordinate health and safety protocols; monitor stressed firms; central bank cuts and use of buffers; temporary regulatory measures; strengthen out-of-court restructuring.
  - Phase 1 examples: Thailand US$700 million for domestic tourism; Fiji US$400 subsidy for first 150,000 tourists; Costa Rica moved holidays to boost weekends; Barbados Welcome Stamp visa.
  - Phase 2 examples: Costa Rica and Jamaica earned WTTC Safe Travels Seal; travel bubbles require collaboration and data sharing.
  - Phase 3: shift from liquidity support to restructuring; job training and targeted fiscal interventions; tailor long-term policy taxonomy.

### Country case summaries (selected excerpts)
- Costa Rica:
  - National parks/protected areas cover about 25 percent of landmass.
  - 2019 international visitors: more than 3 million.
  - Visitor origin shares (2019): North America 53.1 percent; Central America 22.3 percent; Europe 15.9 percent.
  - Tourism supported an estimated 12.9 percent of total employment in 2019.
  - Informal employment in tourism: 59 percent of jobs in 2019.
  - Tourism direct+indirect contribution rose from 12.1 percent of GDP in 2010 to 13.2 percent in 2019.
  - Tourism accounted for about 20 percent of exports.
  - COVID-19 impacts:
    - International tourist arrivals dropped by 98.7 percent year over year in Q3 2020.
    - Direct employment in tourism declined by 28.0 percent year over year (Q3 2020), after a 51.6 percent decline in Q2 2020.
    - Sector economic activity estimated to decline by 40 percent in nominal terms in 2020.
  - Policy responses: 16 health/hygiene protocols; WTTC Safe Travels Seal; mobile app for information and contact tracing; Vamos a Turistear domestic campaign; holidays moved to Mondays by law; international marketing campaigns.
- Jamaica:
  - Tourism and travel represented 31 percent of GDP in 2019; net tourism receipts 56 percent of exports.
  - One-third of tourists older than 50.
  - COVID-19 impacts:
    - Tourist arrivals declined by 66 percent as of end-October 2020.
    - Hotels and restaurants real value added: -14 percent (Q1 2020), -86 percent (Q2 2020), -65.2 percent (Q3 2020).
  - Policy responses: reopened borders June 15, 2020; WTTC Safe Travels stamp; Resilient Corridors; free online training certification for 10,000 tourism workers; Jamaica Cases traveler protection program; temporary cash transfers and grants; bank moratoria on loan repayments.
  - Risks/opportunities: older tourist profile may slow recovery; opportunity to expand community-based tourism (79 percent of US tourists interested).
- Fiji and Vanuatu:
  - 2019 air visitors: Fiji ~900,000; Vanuatu ~120,000.
  - Tourism contributes to more than a third of GDP and employment in both countries; ~70 percent of visitors from Australia and New Zealand.
  - COVID-19 impacts:
    - Borders closed; as of Nov 2020 confirmed cases: Fiji 38; Vanuatu 1.
    - By May 2020, ~70 percent of Vanuatu tourism employees lost jobs.
    - Real GDP estimated to contract by 19 percent (Fiji) and 9.2 percent (Vanuatu) in 2020.
    - No cruise ship landings since March 2020; cruise arrivals 2018: Fiji 188,000; Vanuatu 235,000.
  - Fiscal support:
    - Fiji NPF paid out ~US$24 million (0.5 percent of GDP) by early July.
    - Vanuatu NPF paid out US$12.5 million (1.5 percent of GDP).
    - Fiji tax cuts, transfers, subsidy to Fiji Airways for first 150,000 tourists; Vanuatu reimbursed 30,000 vatu per employee (~US$266) for four months plus tax relief and SME grants.

### Export development, quality ladders, and diversification potential
- Many TDCs have scope to strengthen other sectors; non-tourism exports per capita vary widely.
- Tariff policy: "Most TDCs have scope to lower their relatively high average tariffs (about 10 percent)."
- Example: "In 2019, per capita tourism exports of micro-state St. Kitts and Nevis at about US$7,500 was about three times higher than Malaysia’s manufacturing exports per capita."
- Quality ladders and complexity analyses identify countries with nearby diversification opportunities (e.g., Thailand) and those with limited options (Cambodia, Dominican Republic, Honduras, Jamaica, Nicaragua, Mongolia, Panama).
- Determinants of diversification: proximity to markets, productivity-related variables, unit labor costs, governance, education, infrastructure, tariffs, and labor market rigidity.
- Policy recommendations:
  - Strengthen education and retraining; expand vocational training, especially for women.
  - Upgrade ports and telecommunications; build disaster-resilient infrastructure.
  - Reduce policy bias favoring tourism (tax/tariff exemptions); lower tariffs toward "about 10 percent".
  - Reform labor-market rigidities; enhance regional cooperation for scale and harmonize fiscal incentives.

### Key numeric/modeling parameters and results (selected)
- GIMF calibration highlights:
  - Tourism consumed (% of GDP): Pacific Islands 8.0; Caribbean 5.0; ASEAN-5 6.0; United States 2.0; China 3.6.
  - Tourism produced (% of GDP): Pacific Islands 28.9; Caribbean 18.3; ASEAN-5 7.8; United States 2.8; China 2.0.
  - Share of Global GDP (%, US$): United States 24.23; China 15.74; Europe 21.79; ASEAN-5 3.04; Pacific Islands 0.01; Caribbean 0.15.
- Fiscal stimulus modeling (ASEAN-5 illustrative):
  - Package size: 3 percent of GDP for two years; 1.5 percent of GDP in third year.
  - First-year real GDP loss: 1.6 percent (lower-multiplier); 0.8 percent (higher-multiplier).
  - Cumulative deficit: 7.5 percent of GDP.
  - Debt-to-GDP increases: 6.6 percent (lower-multiplier); 6.1 percent (higher-multiplier).
- Disease elasticity examples (2SLS-IV):
  - SARS, Origin coefficient: -0.387*** [0.092]; scaling informs large estimated COVID-19 impacts (upper-bound 82.5 percent decline).

### Concluding outlook and policy priorities
- Recovery in international travel and tourism will be protracted and highly uncertain.
- Immediate global priorities: establish global safety and health protocols and make a reliable vaccine widely available.
- Fiscal and macro-financial policies can mitigate scarring:
  - Carefully designed fiscal stimulus targeted to the tourism sector, mindful of fiscal space and debt sustainability.
  - Targeted policies for youth and women, including via digitalization.
  - Monetary and financial policies to provide credit relief and early restructuring for distressed but viable firms.
- Long-term priorities: strengthen healthcare systems, shift toward sustainable tourism models, invest in technologies and digital skills, and diversify within and away from tourism to reduce dependence on a single sector.

*IMF Departmental Paper — tfotppwea (source content provided).*

### Executive Summary ������������������������������������������������������������������������������������������������������

### Executive Summary

### The COVID-19 shock and immediate impacts
- The COVID-19 pandemic led to a sudden stop in travel and a collapse in economic activity worldwide.
- Tourism accounts for more than 10 percent of the global economy and is a major exporter of foreign exchange for many countries.
- The pandemic has had severe repercussions on the complex global tourism supply chain, putting millions of tourism jobs at risk.
- Informal and migrant workers, particularly women and youth, have suffered disproportionately from diminished employment opportunities and lack of access to social safety nets, leading to increased poverty and slowing progress toward the UN Sustainable Development Goals.
- Many tourism dependent economies entered the pandemic with limited fiscal space, inadequate external buffers, and foreign exchange revenues extremely concentrated in tourism.
- The paper finds strong correlation between the spread of COVID-19 and big data high-frequency indicators on travel, suggesting healthcare system quality will be pivotal in the post-pandemic recovery of the tourism sector.

### The role of tourism before COVID-19
- Tourism became one of the world’s most important growth engines, accounting for more than 10 percent of global GDP directly and indirectly.
- International tourists rose to more than 1.5 billion by the end of 2019 from 680 million in 2000.
- Spending by international tourists amounted to nearly US$1.6 trillion as of the end of 2019.
- Domestic tourism comprised more than 70 percent of total tourism spending in 2019.
- Domestic tourism spending rose from US$2.2 trillion in 2005 to US$4.5 trillion by 2019; China accounts for over a fifth of such spending.
- On average, tourism directly accounts for about 3½ percent of global GDP; total contribution (direct, indirect, induced) is estimated to exceed 10 percent of global GDP.
- The paper focuses on tourism-dependent countries (TDCs), defined as those where the contribution of tourism and related sectors was above 10 percent of GDP during 2016–18.

### Employment and structural characteristics
- Tourism employed above 300 million globally, providing critical jobs to youth and women and, in several higher-income economies, to a significant share of migrant workers.
- Tourism generates 1 out of 10 youth jobs (WTTC, 2019).
- Women account on average for 54 percent of those employed in the tourism sector compared to 39 percent in the broader economy.
- The sector is characterized by low-skill intensity, high seasonality, high turnover, and a high degree of informality, implying actual employment may be significantly larger than official figures.
- Migrant workers represent a significant share in some countries (example estimates: 16 percent of tourism workers in the European Union and 20 percent in the United States are foreigners; in Thailand, a fifth of hospitality workers come from neighboring countries).
- An ILO survey in ASEAN reveals nearly 97 percent of unemployed migrants in the tourism sector have no access to their host country’s social safety nets (ILO 2020).

### Empirical and modeling findings
- An augmented gravity model drawing lessons from past epidemics finds the impact of infectious diseases on tourism flows is much greater in developing countries than in advanced economies.
- Forward-looking simulations using the IMF’s Global Integrated Monetary and Fiscal model (GIMF) suggest:
  - There is scope for a faster recovery if rapid advancements in vaccine distribution bring travel back to pre-pandemic levels.
  - Significant downside risks exist from protracted uncertainty and limited vaccine effectiveness and availability.
  - Deep and long-term scarring effects could amplify existing vulnerabilities in tourism-dependent economies.

### Structural challenges and opportunities for a “new normal”
- Tourist behavior is likely to shift away from high-density, long-haul, shorter-stay travel, which may raise travel costs and change the rate of return of tourism relative to other sectors.
- Shifts may prompt reallocation of capital and human resources, with broader repercussions for accommodation, retail, and transportation.
- Transitioning from high-density to socially distanced tourism faces inherent rigidities; the protracted process of building stakeholder consensus could amplify pre-pandemic macro and structural vulnerabilities.
- The recovery presents an opportunity to accelerate a shift toward eco-sustainable tourism: lower density, higher value-added, and greater digitalization to reduce health risks, foster a greener recovery, and diversify economies.

### Policy framework: three phases and policy considerations
- Phase 1, crisis mitigation:
  - Many countries provided fiscal support to buttress demand for the industry and preserve jobs.
  - Further support may be needed; there is scope for well designed fiscal stimulus to support the most affected sectors including the poorest households and businesses, while being mindful of available fiscal policy space and debt sustainability concerns.
- Phase 2, reopening:
  - As economies and borders reopen, special attention should be devoted to health and hygiene protocols.
  - Domestic tourism is being incentivized through attractive offers from hotels and tour operators and integrated into governments’ reopening strategies.
  - Creation of COVID-free travel bubbles shows some potential across regions, despite implementation challenges.
  - Targeted policies to address the pandemic impact on youth and women and to enhance access to new opportunities, including through digitalization, can help mitigate scarring, broaden inclusion, and lift potential growth.
  - Monitoring and promoting timely restructuring and retooling is critical as many firms, especially SMEs, risk moving from liquidity stress into insolvency.
- Phase 3, recovery:
  - A shift to eco-sustainable tourism services could reduce health risks associated with mass travel, foster a greener recovery, and help diversify economies.
  - This juncture presents an opportunity to accelerate long-term structural transformation within and beyond tourism to mitigate impacts on output and jobs.

### Global coordination and final conclusions
- Country-specific initiatives will be critical to mitigate scarring effects, but global coordination is necessary given tourism’s worldwide role and the complexity of the global tourism supply chain.
- Immediate global priorities include establishing global safety and health protocols and making a reliable vaccine widely available.

*IMF Departmental Paper — Executive Summary*

### 1. Tourism Share of GDP

### 1. Tourism Share of GDP

### Tourism contribution and employment: selected countries (Asia-Pacific and Western Hemisphere)
- Figure 3 (2018) presents tourism share of GDP and tourism share of employment for selected Asia-Pacific and Western Hemisphere countries. Country lists include: Solomon Islands; Singapore; Mongolia; Australia; China; Sri Lanka; Malaysia; Lao PDR; Tonga; Hong Kong SAR, China; New Zealand; Kiribati; Philippines; Thailand; Cambodia; Fiji; Vanuatu; Macao SAR, China; Maldives; Argentina; El Salvador; Chile; Nicaragua; Costa Rica; Panama; Honduras; Mexico; Dominican Republic; Bermuda; Grenada; St. Vincent and the Grenadines; St. Kitts and Nevis; Jamaica; Dominica; Barbados; Belize; St. Lucia; The Bahamas; Antigua and Barbuda; Aruba.

### Youth and female employment, informality, and wages in tourism (2010–19)
- Share of youth employment in tourism, 2010–18: Figure shows percent scale up to 60 percent for youth share in tourism employment.
- Share of female employment in tourism sector and gender wage gap, 2019:
  - Panel 2 includes countries in the Asia-Pacific and Western Hemisphere regions with a total employment contribution to tourism of above 10 percent.
  - The wage gap is defined as the average wage of women employed in all food and accommodation (as a proxy for the tourism sector) sectors as a share of that of men. Dots below 100 imply women are on average paid less than men.
- Total tourism contribution to employment and share of informal employment, 2019:
  - Informal workers as share of total employment plotted on a 0 to 90 percent scale.
  - Tourism contribution to employment (percent) plotted on a 0 to 25 percent scale.
- Average wage differential in tourism relative to the broader economy: selected countries:
  - Panel 4 refers to the average wage differential for accommodation and food services for tourism relative to the average wage in the economy. Data are for latest available year.
  - Country examples plotted include HND, MNG, MEX, CRI, THA, PAN, KHM, IDN, PHL, DOM, NZL, FJI, LKA, MDV, PRT, ESP, GRC, JAM, COL, CHL, ARG, BRA, ECU, NIC, URY, IND, PNG, ITA, AUS, GRC, BHS, PHL, ESP, FRA (as shown in figure labels).

### MSMEs, sector structure, and resilience
- The tourism sector is characterized by a high share of micro, small, and medium enterprises (MSMEs), often individually or family-owned, particularly in emerging and developing economies, notably in the accommodation, restaurant, and tours and services segments (OECD 2008).
- Country examples of MSME prevalence:
  - In Costa Rica, 94 percent of hotel and lodging firms are classified as micro and small enterprises.
  - In Thailand, SMEs comprise more than 90 percent of firms in the hospitality sector.
- Consequences for resilience:
  - Owing to their size and constrained access to finance—compared, for example, to large multinational groups—these companies have fewer resilience and diversification options to deal with shocks.

### Macroeconomic initial conditions in tourism-dependent economies (2018)
- Figure 5 displays macroeconomic initial conditions in tourism-dependent economies, including:
  - International reserves (months of imports) plotted against tourism contribution to GDP (percent).
  - Public debt (percent of GDP) plotted against tourism contribution to GDP (percent).
- Many tourism-dependent countries entered the pandemic with:
  - Limited fiscal space and weak public sector balance sheets.
  - Inadequate external buffers and foreign exchange revenues highly concentrated in tourism, particularly small states in the Pacific and the Caribbean.
- Structural challenges cited for several countries:
  - Rising inequality.
  - Infrastructure gaps (for example, Cambodia, Lao PDR, Myanmar, and Vietnam).
  - High private debt (Thailand).
  - Limited access to financing for SMEs, pervasive informality, and capacity constraints (small island states).

### COVID-19: shock to tourism and macroeconomic implications
- Containment measures and travel deterrents:
  - Long (usually two-week) quarantine requirements and potential for unexpected health costs deter travelers and make trips longer and costlier.
  - Hotels, retail, indoor gatherings (such as restaurants and entertainment) remain with restrictions or carry a high level of risk.
  - Air travel remains constrained, with global flight numbers declining by 46 percent year over year by October 2020.
  - Entry restrictions persist for countries with high caseloads; the risk of sudden policy changes around visas and quarantine requirements remains elevated.
- Collapse in international tourist arrivals (as of August 2020, UNWTO preliminary data):
  - Fall of 70 percent in international tourist arrivals in the first eight months of 2020, with an 80 percent drop in August, compared to the same period in 2019.
  - UNWTO (2020) suggests a drop in tourist arrivals of 78 percent in 2020 would translate into a loss in visitor spending of US$1.2 trillion, placing more than 100 million direct tourism jobs at risk.
  - Domestic tourism recovered somewhat relative to international tourism; example: as of July 2020 in the European Union, tourist nights by domestic travelers had fallen by about 40 percent compared to 70 percent by nonresidents relative to the previous year.
- Mobility and internet-search indicators:
  - Community mobility average decline relative to pre-pandemic baseline reached as much as 47 percent in Asia-Pacific and 75 percent in the Western Hemisphere (Figure 8, panel 1).
  - Strong correlation between the spread of COVID-19 and the volume of travel-related internet searches; Google Trends data show collapse in travel-related searches after a temporary spike in March, partial recovery during the summer, and renewed depression during the second wave.
- Broader macroeconomic and social challenges:
  - The shock to international travel and tourism is generating significant macroeconomic and social challenges, especially for small states in the Pacific and the Caribbean.

*Source: IMF staff summary of figures and text in the PDF chapter "1. Tourism Share of GDP" from the provided content.*

### 2. COVID-19 and Google Search Trends

### 2. COVID-19 and Google Search Trends

### The Tourism Landscape
- Tourism-dependent economies (TDCs) with a high share of tourism in output and employment, weak external and fiscal positions, and high informality are expected to be the most severely hit.
- Vulnerable groups: women, youth, and migrant workers face disproportionate employment losses and limited access to safety nets; women may drop out of the labor force due to increased domestic responsibilities.
- Remittances: global remittances fell significantly in the second quarter of 2020, with partial recovery in the third quarter; the World Bank estimates that an additional 100 million people could be pushed into extreme poverty this year due to the impact of COVID-19.
- Financial spillovers: liquidity problems in MSMEs and larger service firms (airlines, international hotel and restaurant chains, packaged tour operators) can transform into solvency problems, straining the banking sector—monitoring and timely dealing with stressed or nonperforming loans is critical.
- TDC heterogeneity: countries vary by tourism model (low-cost mass tourism versus full-service tourism) and health capacity (Healthcare Access and Quality Index). Between 2010 and 2015, in a sample of 15 lower- and middle-income countries from the Asia-Pacific and the Western Hemisphere:
  - 3 countries shifted to the right (improved healthcare quality).
  - 1 country shifted toward more full-service tourism.
  - 3 countries moved toward mass tourism.
- Climate risk: global average surface temperature has risen by 1.1 degrees Celsius since 1880, increasing extreme weather and disaster risk; TDCs are especially vulnerable because of reliance on beaches, coral reefs, and specific weather patterns. Cevik and Ghazanchyan (2020) find that a one standard deviation increase in climate change vulnerability is associated with a decline of 9.2 percentage points in international tourism earnings (or a loss of 1.5 percent of GDP) in the Caribbean.

### Lessons from Past Epidemics
- Methodology: an augmented gravity framework using a dataset of 38,184 pairs of countries during 1995–2017 is used to estimate the effect of past infectious-disease episodes on bilateral tourism flows.
- Standard gravity results: higher income levels, proximity/geographical contiguity, cultural similarities (common official language), colonial ties, population, and life expectancy are positively associated with tourist flows.
- Role of infectious diseases:
  - Malaria and yellow fever show smaller estimated coefficients on tourism flows.
  - Ebola and SARS show statistically and economically significant negative effects.
  - Example: for SARS, a 10 percent increase in the number of confirmed cases leads, on average, to a reduction of 4.7 percent in international tourist arrivals.
- Disease characteristics explaining heterogeneity:
  - Vector of transmission: malaria and yellow fever (mosquito-borne) vs. Ebola and SARS (human-to-human), with Ebola and SARS spreading more readily in dense cities and airports.
  - Existence of treatment or vaccine: yellow fever (vaccine) and malaria (treatment) reduce perceived travel risk; Ebola and SARS lacked vaccines/treatments at the time of analysis.
  - Endemic presence vs. temporary outbreak: endemic diseases (malaria, yellow fever) produce different travel responses than temporary outbreaks (Ebola, SARS).
- 2SLS-IV estimates (selected coefficients from Table 1):
  - Ebola: Origin = -0.065*** [0.005]; Destination = -0.089*** [0.008]
  - Malaria: Origin = -0.001 [0.001]; Destination = -0.007 [0.001]
  - SARS: Origin = -0.387*** [0.092]; Destination = -0.078*** [0.104]
  - Yellow fever: Origin = 0.004 [0.012]; Destination = -0.017 [0.104]
  - Controls: Yes; Fixed effects: Yes; Adjusted R2 values shown as 0.83, 0.82, 0.83, 0.82 across specifications.
- Distributional effects:
  - Infectious-disease impacts are significantly higher in developing countries (notably Asia, Latin America, and the Caribbean) than in advanced economies.
  - Example: in the case of SARS, a 10 percent increase in confirmed infections led to a decline of about 8 percent in international tourist arrivals to developing countries—almost twice the average impact on all countries.
- Implications for COVID-19:
  - Similarities between COVID-19 and SARS (both coronaviruses) suggest substantial tourism declines when scaling SARS coefficients to COVID-19 prevalence.
  - Scaling the estimated coefficient of SARS to COVID-19 prevalence yields an approximate decline of 82.5 percent in international tourism flows globally.
  - This upper-bound estimate is broadly consistent with the actual impact of -65 percent registered in the first half of 2020.
  - The paper notes that these estimates should be considered an upper bound because COVID-19 is a global phenomenon and global economic growth is projected to decline by 7 percentage points in 2020 relative to 2019, whereas past epidemics lowered growth by 0.6 percentage points on average.

### Analyzing Macroeconomic Scenarios Under the New Post-Pandemic World
- Modeling approach: a customized version of the IMF’s Global Integrated Monetary and Fiscal model (GIMF) with a tourism module (no direct epidemiological dimension) is used to construct scenarios.
- Scenario design:
  - Benchmark scenario: near-total shutdown of tourism followed by illustrative gradual recovery, hampered by permanent economic scarring; international tourism begins to resume in the second year and is expected to return near the pre-pandemic level after three years, reaching its previous growth path a few years later (aligned with October 2020 WEO projections). Benchmark assumes relatively equitable and consistent vaccine distribution restoring tourist confidence over several years.
  - Alternative downside and upside scenarios are used to highlight uncertainty; Chapter 4 presents two policy scenarios showing how governments could influence recovery.
- Economic scarring assumptions:
  - Productivity in the tourism sector is assumed to be 10 percent lower than before for a given allocation of capital and labor, capturing additional costs (health/cleaning protocols, social distancing, contactless measures).
  - Supporting evidence: World Bank (Dieppe, Kilik Celik, and Okou 2020) suggests productivity after epidemics has declined historically by 6 to 15 percent.
  - Theoretical models (Kozlowski, Veldkamp, and Venkateswaran 2020) indicate large pandemics can produce long-term persistent productivity decline.
- Regional impacts (benchmark GIMF simulation):
  - Regions with larger tourism sectors experience the largest real GDP losses; the Pacific Islands and the Caribbean are the hardest hit, with notable impacts in other major tourism regions.
  - Some regions that are net importers of tourism services (Latin America and China in particular) may see increases in real GDP despite tourism shocks.
- Policy-relevant points emphasized in the analysis:
  - Rapid, equitable global deployment of vaccines and treatments is crucial to restore tourist confidence.
  - Policy interventions that cushion income losses and address long-term scarring can materially affect recovery trajectories.
  - Monitoring and addressing stressed or nonperforming loans in banking systems that entered the crisis with weak balance sheets is critical to recovery.

*Source: IMF staff calculations and analysis as presented in the chapter "2. COVID-19 and Google Search Trends" of the PDF content unit.*

### 1. Pacific Islands

### 1. Pacific Islands

### Overview of impacts on the Pacific Islands
- Real GDP falls by more than 20 percent in tourism-dependent regions (Pacific Islands and Caribbean).
- The tourism collapse leads to large falls in employment, reducing household income and hence consumption, because most households are not able to smooth consumption.
- Investment, mainly related to tourism, also contracts.
- Because the tourism sector is relatively larger in the Pacific Islands, there is a larger wealth shock, leading to a significantly larger depreciation in the real effective exchange rate (REER).
- The Pacific Islands are expected to have a stronger net export response, with lower imports and higher exports resulting from the shift in the REER.
- Comparing only real GDP for economies such as the Pacific Islands and the Caribbean masks different impacts between the domestic and external sectors.

### Fiscal positions, debt dynamics, and financing constraints
- Many governments in the Pacific have accumulated significant debt, whose burden is exacerbated by the COVID-19 shock and weighs on the recovery.
- The benchmark scenario assumes a deterioration in countries’ fiscal positions and a sharp increase in debt, due to growth and exchange rate shocks.
- It is assumed that government discretionary spending is unchanged from pre-pandemic budget plans; both regions’ governments provide small automatic stabilizers.
- The share of government spending as a percent of GDP is higher in the Pacific Islands than in the Caribbean, so as real GDP recovers, the debt-to-GDP ratio returns faster to its pre-pandemic values in the Caribbean than in the Pacific Islands.
- The benchmark assumes all debt can be financed easily at the (low) global interest rate; however, limited market access in many countries makes a mix of financing, including better terms from development partners and policy adjustment, more likely.
- If affordable financing were to be replaced by cuts in government consumption or social transfers:
  - the fiscal position would not improve;
  - private consumption would be substantially weaker;
  - negative impacts on real GDP would increase the debt-to-GDP ratio.

### Scenario design and key parameter differences
- Main source of uncertainty: length of the pandemic and path of recovery; global trough for tourism sector is in 2020, with alternative scenarios diverging from the benchmark as recovery begins in 2021.
- Downside scenario:
  - Slower recovery for tourism demand.
  - Differentiation by tourists toward regions perceived to have stronger healthcare systems.
  - Sectoral productivity reduction of 20 percent (instead of 10 percent in the benchmark).
  - Larger permanent loss of GDP and slower recovery in employment, although employment eventually returns to former levels.
  - Employment may overshoot in the short term as firms hire labor to move to profitability while investment takes time to rebuild capital.
- Benchmark scenario:
  - Tourism sector productivity shock of 10 percent.
- Upside scenario:
  - Faster recovery for tourism demand and supply.
  - Temporary surge in international tourism in 2021 due to pent-up demand if vaccine rollout and herd immunity permit.
  - No productivity shock to the tourism sector (no economic scarring).
  - Real GDP and employment recover more rapidly and to higher levels than in the benchmark.
  - For Pacific Islands and Caribbean, measured real GDP still recovers less than other countries, though consumption, investment, exports, and real income fully recover in the upside scenario.

### Modelled outcomes and behavioral notes
- Employment metric includes hours worked; increases could reflect more hours rather than higher employment headcount.
- In the downside scenario, despite slower investment, firms may temporarily increase hours worked or hire to restore profitability, producing short-term employment overshoots.
- The Pacific Islands display a pronounced REER depreciation in response to the tourism shock, analogous to an oil exporter facing a large fall in global oil demand.
- The Pacific Islands’ larger tourism share produces a larger wealth shock and a stronger net export response relative to the Caribbean.

### Policy options and sequencing for tourism-dependent countries (TDCs)
- Immediate priorities: coordinate public and private sector to establish and communicate health and safety protocols; coordinate aviation, hospitality, and insurance sectors.
- Monitoring and supporting stressed firms early is critical to prevent liquidity stress turning into bankruptcies and macro-financial spillovers.
- Financial sector responses:
  - Central bank policy rate cuts and use of capital and liquidity buffers can create a favorable environment for borrowers.
  - Financial authorities can provide temporary regulatory and supervisory measures.
  - Caution advised against relaxing loan classification, provisioning rules, or debt restructuring processes.
  - Banks can facilitate debt restructuring for viable firms, possibly with central bank funding support.
  - Strengthened legal frameworks, including out-of-court procedures, can speed restructuring of viable tourism SMEs.
- Phase 1 (crisis-mitigation):
  - Tourism-specific fiscal and financial measures for businesses and workers, national airlines, and SMEs.
  - In diversified economies with fiscal space, consider broader stimulus to activate the economy.
  - Examples of national measures:
    - Thailand allocated US$700 million to spur domestic tourism.
    - Fiji introduced a subsidy of US$400 for the first 150,000 tourists.
    - Costa Rica moved several national holidays to Mondays to extend weekends in 2020 and 2021 to boost domestic tourism.
    - Barbados’ Welcome Stamp visa (one-year residency for remote employees) generated significant demand.
- Phase 2 (reopening):
  - Emphasize health and hygiene protocols and certified hygiene standards to restore traveler confidence.
  - Costa Rica and Jamaica introduced strict health and hygiene protocols and earned the World Travel and Tourism Council’s Safe Travels Seal.
  - Thailand uses strict quarantine and limited travel programs.
  - Capacity constraints in healthcare systems in many small states may complicate adoption of safety protocols and delay reopening.
  - Travel bubbles between low-incidence regions can facilitate gradual reopening but require collaboration among governments, airports, and airlines and real-time data sharing.
- Phase 3 (recovery):
  - Transition from exceptional liquidity support to corporate restructuring of viable firms.
  - Calibrate policies to country heterogeneity; consider targeted fiscal interventions, job training, and support for transitions to new sectors such as digital services.
  - Long-term policy taxonomy to address permanent scarring should be granular and country-specific.

### Illustrative fiscal stimulus evidence (ASEAN-5 application) and relevance to Pacific Islands
- Two versions of a three-year fiscal stimulus package modelled:
  - Package size: 3 percent of GDP for two years and 1.5 percent of GDP in the third year.
  - Lower multiplier package focuses on public consumption, transfers to all households, and cuts in labor income taxes.
  - Higher multiplier package focuses on infrastructure investment, transfers targeted to poorer, liquidity-constrained households, and a cut in the value-added consumption tax (VAT).
- First-year real GDP loss under the fiscal packages:
  - 1.6 percent under the lower multiplier package.
  - 0.8 percent under the higher multiplier package.
- Fiscal costs and debt implications:
  - Cumulative deficit of 7.5 percent of GDP.
  - Debt-to-GDP ratio increases by:
    - 6.6 percent of GDP under the lower multiplier package.
    - 6.1 percent of GDP under the higher multiplier package.
- Policy implications for Pacific Islands:
  - Many Pacific Island states have limited fiscal space and may not be able to implement large stimulus packages.
  - Development partner financing and tourism-specific fiscal support may be necessary in small TDCs.

### Travel bubble scenario (Australia, New Zealand, Pacific Islands)
- Counterfactual worse scenario: tourism shut down for the first two years instead of only the first year.
- Alternative: pursue a travel bubble in the second year, before general reopening in the third year.
- Under the travel bubble:
  - Regions reopen tourism internally and to each other in Year 2.
  - Pacific Islands gain very large GDP gains due to the outsize role of tourism.
  - Australia and New Zealand see more modest GDP gains, mostly from reviving investment.
  - Pacific Islands benefit more in tourism exports as demand from Australia and New Zealand is greater than pre-COVID-19.
  - Tourism exports by Australia and New Zealand barely move because Pacific Islands send relatively few tourists due to income levels and small size.
  - Travel bubble reduces the risk of COVID-19 cases caused by tourists from outside the bubble (benefit not quantified in the model).

*Source: IMF staff calculations.*

### Box 2. Implementing a Travel Bubble

### Box 2. Implementing a Travel Bubble

### Impact Illustration
- Box Figure 2.1: "Impact of a Travel Bubble on Australia, New Zealand, and the Pacific Islands" — figure shows percent deviations from the pre-COVID-19 forecast for:  
  - Real GDP (percent difference)  
  - Tourism (percent difference)  
  - Tourism Exports (percent difference)  
  - REER (Percent difference; +ve is an appreciation)  
- Figure note: "No tourism in years 1 and 2" and "REER = real effective exchange rate."  
- Source for calculations stated as "IMF staff calculations."

### Strengthening health systems and health tourism
- Strengthen TDCs’ health systems to enhance attractiveness post-COVID-19 and to promote a health tourism industry in some countries.  
- Continued emphasis on early detection, treatments, and social distancing to improve COVID-19 health outcomes.  
- A global solution to facilitate and accelerate vaccine production and distribution as a global public good would:  
  - facilitate complementary macro and financial policies discussed in the paper; and  
  - minimize risks of crowding out small, highly tourism-based economies with weak healthcare systems that could face delayed vaccine access and limited storage capabilities.  
- For countries with relatively weak healthcare infrastructure, governments should support:  
  - building, upgrading and accreditation of healthcare facilities;  
  - capacity building of medical personnel and support staff; and  
  - measures to ensure availability and access to quality medicines and diagnostic equipment.  
- Accreditation of health facilities by reputable institutions is important to attract visitors for medical treatments.  
- Political stability, infrastructure enhancement, and medical education are important complements.

### Low-density and eco-sustainable tourism
- Promote low-density and eco-sustainable tourism services to reduce health risks.  
- Transition requires a coherent, multi-sectoral strategy involving industry, private sector, government, and civil society and embedding plans in national development frameworks.  
- Eco-tourism is highlighted as fast-growing and higher value-added, with demand from high-income foreigners likely to receive vaccine early.  
- Costa Rica example: focus on eco-sustainability and reforestation, extensive marketing, infrastructure to access ecotourism sites, business development advisory, cooperation with international researchers.  
- Thailand example: shift from mass tourism to niche, higher-value added tourism to move up the value chain and reduce carbon footprint and damage to natural resources.  
- Mass tourism need not stop entirely, but all countries must explore ways to manage tourism more sustainably and promote social distancing.  
- Political stability and improvements in infrastructure and human capital help destinations stand out.

### Technological innovation and digitalization
- Technology can ensure safety and protection of tourists and workers, enhance access to information, and enable rapid response to disruptions.  
- Uses include sharing information, enhancing contact tracing, and supporting touchless service delivery (for example, digital concierge).  
- Technology facilitates digitally self-guided tourism consistent with social distancing.  
- SMEs must accelerate integrating digital capabilities to avoid widening the digital divide.  
- Automation will change job composition—some jobs obsolete, others transformed, and new opportunities created.  
  - Reskilling and training tourism employees to interact with digital resources and adapt to new customer-service requirements is critical.  
  - Digitalization can broaden inclusion, for example through telework, facilitating greater access to opportunities for women.  
- The impact of automation on jobs depends on the balance of substitution and enhancement effects (Ivanov (2020) referenced in text).

### Scope for strengthening other sectors and export diversification
- TDCs can offset a permanent decline in arrivals by strengthening other sectors; many TDCs already have significant non-tourism exports per capita (Figure 18 referenced).  
- Natural resources abundant in some (Chile, Mexico) could allow expansion of non-tourism exports where reserves permit.  
- Some small states rely on passport sales and financial services benefiting from low taxation, which are subject to risks.  
- Limits to diversification: many small TDCs’ tourism exports per capita are multiple times higher than manufacturing exports in East Asian exporters; structural adjustment would be significant.  
- Example numerical points preserved exactly from text:  
  - "Most TDCs have scope to lower their relatively high average tariffs (about 10 percent)."  
  - "In 2019, per capita tourism exports of micro-state St. Kitts and Nevis at about US$7,500 was about three times higher than Malaysia’s manufacturing exports per capita." (Salinas (forthcoming-b) cited in text.)  
  - "If ECCU micro-states would increase their manufacturing exports per capita to the level of St. Kitts and Nevis, it would allow them to offset more than a 10 percent permanent decline in their tourism exports."  
- Strengthen tourism linkages to agriculture, manufacturing, and entertainment to increase domestic value added; example: Jamaica’s Agri-linkages Exchange (ALEX) online platform to match hotel buyers with local farmers.  
- Quality upgrades and expansion: many TDCs have scope for quality upgrades across sectors including food and live animals, crude materials, machinery, and manufacturing goods (Figure 19 and Annex Figure 4.4 referenced).  
- Product complexity constraints: several TDCs (Cambodia, Dominican Republic, Honduras, Jamaica, Nicaragua, Mongolia, Panama) lack very complex export baskets and have limited options to move to more complex products. Jamaica’s highest potential opportunities noted in industrial machinery products and plastics.  
- Development of new products fosters innovation and human capital but is challenging for small states; countries tend to move to products requiring similar know-how (Hausmann and others (2014) cited).

### Policy recommendations to enhance exports and resilience
- Strengthen education and retraining programs to allow swifter sector reallocation; expand vocational training, particularly to benefit women and strengthen female labor force participation.  
- Upgrade ports and telecommunications infrastructure; telecommunications key to nurturing the digital economy.  
  - Build infrastructure resilient to natural disasters (noted as particularly important).  
- Eliminate policy bias unduly favoring tourism relative to other sectors; many TDCs can lower average tariffs "about 10 percent" and reduce tax and tariff biases that favor tourism.  
- Reduce labor market rigidities to strengthen competitiveness and enable sector reallocation; flexible labor policies can help ensure competitive unit labor costs, especially in Caribbean countries with strong unionization.  
- Cooperation among regional peers to address low economies of scale and harmonize fiscal incentives; examples of regional integration efforts referenced:  
  - CARICOM’s Single Market and Economy established in 2001.  
  - PACER Plus opened for signatures in June 2017, and entered into force in December 2020 with Australia, Cook Islands, Kiribati, New Zealand, Niue, Samoa, Solomon Islands, and Tonga.  
- Broad, horizontal policies and institutional choices can foster export development and also support tourism exports; many of these require public investment and support from larger countries and development partners.

### Concluding outlook
- Recovery in international travel and tourism will be protracted and subject to high uncertainty.  
- Past infectious diseases had significant negative effects on international tourism flows; COVID-19’s effect is much greater due to high contagiousness and global spread.  
- Model-based scenario analysis indicates a gradual recovery hampered by permanent economic scarring after a near-total shutdown.  
- Upside exists from large pent-up demand and rapid vaccine rollout that could lead to a faster return of mobility and minimize long-term scarring.  
- Fiscal and macro-financial policies can mitigate deep and long-lasting scarring:  
  - Carefully designed fiscal stimulus targeted to help the tourism sector, mindful of fiscal space and debt sustainability.  
  - Targeted policies for youth and women, including through digitalization, to broaden inclusion.  
  - Monetary and financial policies to provide credit relief and early interventions and restructuring for distressed but viable firms to mitigate macro-financial spillovers.  
  - Interventions should be aimed at solvent strategic firms facing liquidity challenges, given limited fiscal resources and the need to avoid hindering resource allocation.

*Source: Box 2. Implementing a Travel Bubble, tfotppwea - Box 2. Implementing a Travel Bubble.*

### Conclusion

### tfotppwea - Conclusion

### Main conclusions
- The COVID-19 pandemic has inflicted pandemic-related difficulties and economic scarring on tourism-dependent economies, requiring exploration of opportunities for regulatory reforms to mitigate the scarring effect.
- Countries will need to rethink the tourism model and create opportunities for diversification within and beyond the industry through policy support and structural reforms.
- No one-size-fits-all solution can rejuvenate the tourism sector; the new normal for international tourism will differ across and within countries.
- The paper provides a policy taxonomy for long-term sustainability to embrace granular and idiosyncratic elements of the transition.
- Pace and scope of recovery depend on global developments, but policy and institutional choices play a critical role in shaping post-pandemic recovery.
- Beyond immediate mitigation, the focus needs to be on long-term policy solutions: strengthening healthcare systems, shifting to sustainable tourism models, investing in new technologies, and diversifying within and away from tourism to avoid dependence on a single sector.

### Global cooperation
- Sustainable and broad-based recovery in international travel and tourism requires global cooperation.
- Immediate global priority: produce, purchase, and distribute medical treatments and vaccines to halt COVID-19.
- The pandemic presents an opportunity to explore long-term solutions and to restructure and rebuild tourism for more resilient and environment-friendly economies.
- Global cooperation and guidelines on health and safety protocols, and secure platforms unifying individuals, governments, and the travel industry to share information, would provide tourists with good-practice guides and travel requirements (for example, a negative viral test and health insurance coverage).

### Costa Rica: Key findings and policy responses
Background and economic importance
- National parks and protected areas cover about 25 percent of the national landmass.
- In 2019 Costa Rica attracted more than 3 million international visitors.
- Visitor origin shares (2019): North America 53.1 percent; Central America 22.3 percent; Europe 15.9 percent.
- Purpose shares (2016 survey): Leisure 66 percent; Tourism for health 13.4 percent; Business 12.7 percent; Education 7.7 percent.
- Tourism supported an estimated 12.9 percent of total employment in Costa Rica in 2019.
- Informal employment in the tourism sector accounted for 59 percent of all jobs in 2019 (compared to about 46 percent in other sectors).
- More than half of the sector’s employees are women.
- Tourism’s direct and indirect contribution to GDP rose from 12.1 percent of GDP in 2010 to an estimated 13.2 percent in 2019.
- Tourism accounted for about 20 percent of the country’s exports.

Impact of COVID-19
- International tourist arrivals dropped by 98.7 percent year over year in the third quarter of 2020.
- Direct employment in the tourism sector declined by 28.0 percent year over year (up from a decline of 51.6 percent year over year in the second quarter 2020), driven by restaurants and hotels.
- Overall economic activity in the tourism sector is estimated to decline by 40 percent in nominal terms in 2020.
- Following border closure in March 2020, Costa Rica gradually relaxed entry restrictions since August 2020; as of November 2020 tourists from 44 countries, including from the United States, have been authorized to visit.
- Entry requirements include health insurance coverage from local or international providers; quarantine restrictions were removed to support tourism.

Policy responses and initiatives
- The Costa Rican Tourism Institute (ICT), with the private sector, launched 16 health and hygiene protocols for tourism-related activities.
- Adopting these protocols earned Costa Rica the World Travel and Tourism Council’s Safe Travels Seal in July.
- ICT provided tourists with good-practice guides and a mobile app to share relevant information and enhance contact tracing on a voluntary basis.
- Domestic campaign: Vamos a Turistear (let’s go sightseeing) to incentivize domestic tourism with offers, discounts, and financing options until the end of December 2020.
- Congress approved a law moving several national holidays to Mondays to extend weekends during 2020–24 to boost domestic tourism.
- International campaigns: Only the Essentials (United States and Canada) and Un Sanctuaire de Vie (France).
- Additional measures: services for small groups with mandatory local guide and focus on sustainable tourism; attract foreign direct investment; transform the Guanacaste area into a tourism hub; promote maritime tourism.

### Jamaica: Key findings and policy responses
Background and economic importance
- Nearly 80 percent of foreign nationals visiting Jamaica are leisure and holiday travelers.
- All-inclusive hotels represent 60 percent of the total accommodation infrastructure.
- One-third of Jamaica’s tourists are older than 50.
- Stopover visitors and average per-person daily expenditure grew by an annual average of 3.9 percent and 3 percent, respectively, over the last 20 years.
- Tourism and travel represented 31 percent of GDP in 2019.
- Net tourism receipts account for 56 percent of total exports.
- The share of employment in sectors with direct or indirect linkages to tourism is more than 30 percent.
- Hotels and restaurants, 8.4 percent; Transport, 3.0 percent; Wholesale and retail, 19.4 percent; Agriculture, 15.5 percent; Construction, 8.6 percent; Manufacturing, 3.1 percent; Other, 42.1 percent (average share of employment by sector, 2012–20).

Impact of COVID-19
- As of the end of October 2020, tourist arrivals declined by 66 percent.
- Real value added of hotels and restaurants declined by 14 percent in the first quarter of 2020, by 86 percent in the second quarter of 2020, and by 65.2 percent in the third quarter of 2020.
- Output of recreational activities and transport sectors also contracted sharply due to the drop in foreign tourist arrivals and reduction in air flights and cruise passengers.
- Hotel closures negatively affected agricultural sector sales (coffee and cocoa), food processing, manufacturing, and air transport.

Policy responses and initiatives
- Government of Jamaica reopened borders to all international travelers on June 15, 2020.
- COVID-19 travel protocols include prior-to-departure travel authorization, risk-based quarantines, and movement limitations; protocols received WTTC Safe Travels stamp.
- Two Resilient Corridors established where tourism operators trained and certified for adherence to COVID−19 protocols; low-risk tourists are required to stay within these corridors.
- Jamaica Hotel and Tourism Association launched community training; JAPEX was staged virtually with record participants.
- Authorities provided free online training certification classes for 10,000 tourism workers.
- In November 2020 authorities introduced Jamaica Cases traveler protection and emergency service program (field rescue and repatriation for medical emergencies, including COVID-19 and natural disasters).
- Policies to avoid bankruptcies and sustain jobs: temporary cash transfers, safety nets for lower-earning dismissed workers, grants for small tourism operators and informal businesses; Rediscover Jamaica campaign to encourage discounted domestic travel.
- Commercial banks provided temporary moratoria on loan repayments to tourism operators.

Risks and opportunities
- Older-age profile of tourists (one-third older than 50) could result in slower recovery in arrivals.
- Structural constraints: concentration of room capacity in large hotels, high crime, weak infrastructure outside resorts, and low-price competitiveness.
- Domestic spending contributed to only 21 percent of total tourism and travel spending in 2019.
- Opportunity: 79 percent of US tourists interested in community-based tourism (CBT); authorities announced plan to encourage CBT and established Community Tourism Unit within the Ministry of Tourism.
- CBT could complement beach tourism, increase resilience, and generate broader employment, but requires addressing crime and road and water infrastructure.

### Fiji and Vanuatu: Key findings and policy responses
Background and economic importance
- In 2019 Fiji received about 900,000 visitors by air; Vanuatu received about 120,000 visitors by air.
- Tourism contributes to more than a third of GDP and employment in both countries.
- About 70 percent of visitors to both countries come from Australia and New Zealand.
- Fiji: tourism concentrated on Viti Levu and Vanua Levu (333 islands total).
- Vanuatu: tourism concentrated on Efate (Port Vila) and islands such as Tanna, Espiritu Santo, Pentecost, and Aneitym (83 islands total).
- Both countries are highly vulnerable to shocks including tropical cyclones (TC Evan in 2012–13, TC Pam in 2015, and TC Harold in April 2020).

Impact of COVID-19
- Both countries enacted strict border controls in late March and remain closed to most international travel.
- As of November 2020, confirmed COVID-19 cases: Fiji 38; Vanuatu 1.
- Since April 2020, Fiji’s tourism revenue has been minimal; Vanuatu received no revenue from foreign visitors.
- By May 2020, an estimated 70 percent of employees in Vanuatu’s tourism sector had lost their jobs.
- Real GDP is estimated to have contracted by 19 percent (Fiji) and 9.2 percent (Vanuatu) in 2020.
- Cruise ship importance (2018): Fiji received 188,000 cruise arrivals; Vanuatu received 235,000 cruise arrivals.
- Average passenger spending per cruise passenger: Fiji US$44; Vanuatu US$85.
- No cruise ship landings since March 2020, complicating recovery prospects for small businesses dependent on cruise visitors.

*Italic: IMF staff summary of the Conclusion and Annex 1 case studies from the provided document.*

### Annex Figure 1.6. Tourism Contribution to the Economy

### Annex Figure 1.6. Tourism Contribution to the Economy

### Fiji and Vanuatu: Impact and Policy Responses
- Fiscal support measures implemented to support tourism industries:
  - Fiji: tax cuts, transfer payments, and a subsidy to Fiji Airways to incentivize the first 150,000 tourists in the new fiscal year.
  - Vanuatu: reimbursing registered employers 30,000 vatu per employee (US$266) for four months to help retain their workforce, along with business tax relief and SME grants.
- Use of national pension funds to support affected households:
  - Fiji National Provident Fund (NPF) paid out an estimated US$24 million (0.5 percent of GDP) to its members by early July.
  - Vanuatu NPF paid out US$12.5 million (1.5 percent of GDP).
- Constraints and outcomes:
  - Borders largely closed and limited fiscal space make business survival and household support increasingly challenging.
  - Both countries expressed interest in joining travel bubbles (Trans-Tasman, Bula bubble, Tamtam bubble), but none materialized due to local outbreaks in key source countries.
  - Fiji launched a Blue Lanes initiative in August 2020 allowing yachts to berth after strict quarantine and testing; Fiji has received more than 90 yachts through this initiative.
- Tourism positioning and challenges:
  - Both countries could benefit from diversion of visitors staying within the region (Australia, New Zealand) given direct flights mostly within four hours.
  - Potential to offer socially distanced vacations and expand ecotourism, with Vanuatu offering low-density tourism and off-the-beaten-path experiences.
  - Challenges: remoteness, relatively weak health infrastructure, heavily impacted national airlines, disrupted international air and cruise operations.
  - Global Health Security Index: Fiji and Vanuatu rank in bottom quartile, which could deter some future visitors.
- Chart/figure numeric labels present in source (preserved as they appear):
  - 30, 36, 42, 40, 32, 34, 38, 600, 750, 900, 850, 650, 700, 800, 80, 100, 105, 90, 85, 95, 110, 120, 115, 0, 20, 10, 30, 40, 60, 50, 2010121314151617181120101214161113151718
  - 60, –100, –20, 20, –80, –60, –40, 0, 40, –120
  - Sources: UN World Tourism Organization, Yearbook of Tourism Statistics; Country authorities; and IMF staff estimates.

### Thailand: Background, COVID-19 Impact, and Policy Responses
- Pre-COVID-19 importance and structure of tourism:
  - Thailand ranked ninth for tourist arrivals and fourth for tourism receipts in 2018 (UNWTO 2019).
  - At the end of 2019, tourism comprised 12 percent of Thailand’s GDP, and nearly a fifth of the economy once accounting for related services.
  - Tourism employment: about 15 percent of Thailand’s total employment.
  - Migrant workers from neighboring countries estimated at 5 percent of total workers in the tourism industry.
  - Gender composition and wage gaps:
    - Women comprise 65 percent of employment in the accommodation and food services sector.
    - Women earn about 80 percent of men’s wages in the sector.
    - In the rest of the economy, women’s wages are 99.5 percent of that of men.
  - Nearly 40 million tourists visited Thailand in 2019.
- COVID-19 impact on tourism and labor markets:
  - Tourist arrivals expected to decline to just 6.7 million tourists by the end of 2020.
  - Labor dislocation: 139,000 job losses related to the tourism sector in the first quarter of 2020.
  - Thai Chamber of Commerce estimate: up to 6 million could lose employment by the end of the year.
  - Informal and migrant workers, particularly women, likely to suffer disproportionately due to limited access to safety nets.
  - Financial strain on airlines including Thai Airways and several budget airlines.
  - Infections plateaued by April 2020, foot traffic recovered per Google mobility indicators, but a second wave materialized in late December 2020.
- Policy responses and support measures:
  - Soft financing program: US$3 million provided to tour operators as part of COVID-19 fiscal package.
  - Domestic tourism programs approved in June 2020:
    - Three programs worth US$700 million (about 0.14 percent of GDP) to support domestic tourism.
    - Programs subsidized travel for healthcare workers, subsidized accommodation/food/other travel expenditures for qualifying domestic tourists, and subsidized transportation costs for long distance domestic trips.
    - Incentives covered domestic travel undertaken between July and October.
  - Phased reopening strategies for international tourists:
    - Travel bubbles considered but postponed after second wave outbreaks in candidate countries; travel insurance policy design remains elusive.
    - Thailand Special Tourist Visa (STV) introduced in October 2020 for long-stay visitors from low COVID-19-incidence countries with medical insurance and proof of accommodation; 14-day quarantine at certified facilities, negative COVID-19 test upon completion, and mobile phone app tracking required.
      - STV valid for stays up to 90 days and renewable twice.
      - As of the end of November 2020, the scheme drew 825 visitors from 29 different countries (mainly China).
      - In early December 2020, the authorities approved expansion of the STV to visitors from every country.
  - Mixed use of resources: some hotels converted to quarantine facilities and accommodation for medical workers, though occupancy rates remain historically low.
- Strategic long-term adjustments being explored:
  - Shift from mass tourism to low-density high-end tourism to promote more sustainable tourism with a lower ecological footprint.
  - Strengthening healthcare system for rapid prevention, detection, and treatment including access for tourists.
  - Investments in digital/mobile resources and connectivity to support touchless service delivery and digitally self-guided tourism.
- Chart/figure numeric labels preserved as they appear:
  - 1997992001030507091113151719
  - Jan. 2019, Feb. 19, Mar. 19, Apr. 19, May 19, June 19, July 19, Aug. 19, Sep. 19, Oct. 19, Nov. 19, Dec. 19, Jan. 20, Feb. 20, Mar. 20, Apr. 20, May 20, June 20, July 20, Aug. 20, Sep. 20, Oct. 20

### Applying a Gravity Model to Predict Post-Pandemic Tourism Flows
- Literature and motivation:
  - Gravity model widely used for trade, capital movements, migration, and tourism flows.
  - Scarce research modeling bilateral tourist movements in a gravity framework accounting for infectious diseases.
  - Evidence that infectious diseases have a significant negative effect on international tourism flows.
- Data and sample:
  - Unbalanced panel of annual observations for 38,184 pairs of countries during the period 1995–2017.
  - Bilateral tourism flows for 172 countries of origin and 222 countries of destination from the WTO database, yielding a data set of more than 261,488 observations.
  - Main explanatory variable: number of confirmed infectious-disease cases (Ebola, malaria, SARS, yellow fever) obtained from WHO database.
  - Control variables: real GDP, population, real effective exchange rate (REER) from IMF WEO and World Bank WDI; standard gravity variables from CEPII (distance, common official language, colonial history, geographical contiguity).
- Model specification (preserved notation and form):
  - Baseline gravity: ln(Tijt) = b + a ln(GDPit) + g ln(GDPjt) + θ ln(Distij) + ηi + φj + μt + εijt
  - Augmented gravity with infectious diseases: ln(Tijt) = b + a ln(GDPit) + g ln(GDPjt) + θ ln(Distij) + δXijt + ω ln(Virijt) + ηi + φj + μt + εijt
    - Xijt includes log population (origin and destination), REER in destination, binary variables for common language/colonial history/contiguity, life expectancy, government effectiveness in destination.
    - Virijt denotes number of confirmed cases of Ebola, malaria, SARS, and yellow fever scaled by population in origin and destination countries.
  - Robust standard errors clustered at the country-pair level (results broadly unchanged when clustered at the country level).
- Forecasting and evaluation:
  - Out-of-sample forecasting evaluation partitioned sample: estimation sample 1995–2014; forecasting sample 2015–17.
  - Forecast accuracy metrics: MAE, RMSE, U-Theil (formulas preserved as presented).
  - Including infectious-disease episodes in models lowers the RMSE of bilateral tourism flow forecasts by as much as 7 percent compared to the standard model without the number of infectious-disease cases.
- Additional methodological notes preserved:
  - Geographic distance measured as great-circle distance in kilometers between capital cities.
  - Binary variables assigned value 1 if share common official language, colonial tie, or adjacent border; 0 otherwise.

*tfotppwea - Annex Figure 1.6. Tourism Contribution to the Economy*

### Annex 2. Applying A Gravity Model to Predict Post-Pandemic Tourism Flows

### Annex 2. Applying A Gravity Model to Predict Post-Pandemic Tourism Flows

### Empirical gravity-model estimations (PPML)
- Dependent variable: Bilateral tourism flows (in log form).
- Key estimated coefficients (columns [1]–[5]):
  - Real GDP, origin: 0.129***, 0.127***, 0.127***, 0.127***, 0.131*** ([0.006], [0.006], [0.006], [0.006], [0.006])
  - Real GDP, destination: 0.129***, 0.129***, 0.129***, 0.129***, 0.132*** ([0.007], [0.008], [0.008], [0.008], [0.008])
  - Distance: 0.228***, 0.232***, 0.231***, 0.232***, 0.229*** ([0.004] each)
  - Common language: 0.176***, 0.174***, 0.174***, 0.174***, 0.180*** ([0.007] each)
  - Colonial history: 0.080***, 0.083***, 0.083***, 0.083**, 0.076*** ([0.016] each)
  - Geographical contiguity: 0.034, 0.035, 0.035, 0.035, 0.031 ([0.019], [0.016], [0.016], [0.016], [0.016])
  - Population, origin: 0.063***, 0.065***, 0.059***, 0.063***, 0.066*** ([0.011] each)
  - Population, destination: 0.080***, 0.078***, 0.082***, 0.078***, 0.084*** ([0.012], [0.012], [0.012], [0.013], [0.012])
  - REER, destination: 0.000*** (all columns) ([0.000] each)
  - Life expectancy, destination: 0.097, 0.103, 0.113, 0.091, 0.088 ([0.043], [0.043], [0.044], [0.044], [0.043])
- Selected disease variables (where included):
  - Ebola, Origin: 0.013*** ([0.001])
  - Ebola, Destination: 0.010*** ([0.002])
  - SARS, Origin: 0.003*** ([0.02])
  - SARS, Destination: 0.040*** ([0.02])
  - Malaria and Yellow fever coefficients reported as small/insignificant in these specifications.
- Sample sizes and fit:
  - Number of observations: 224,019; 215,589; 215,589; 215,589; 219,132 (by column)
  - Origin FE: Yes; Destination FE: Yes; Year FE: Yes
  - Adjusted R2: 0.77, 0.84, 0.80, 0.83, 0.80 (by column)
- Note: Robust standard errors clustered at country level. *, **, and *** denote significance at the 10%, 5%, and 1% levels.

### Empirical gravity-model estimations (2SLS-IV)
- Dependent variable: Bilateral tourism flows (in log form).
- Key estimated coefficients (columns [1]–[4]):
  - Real GDP, origin: 0.933***, 0.935***, 0.959***, 0.957*** ([0.012], [0.012], [0.014], [0.012])
  - Real GDP, destination: 0.865***, 0.856***, 0.861***, 0.896*** ([0.013], [0.013], [0.015], [0.013])
  - Distance: 1.717*** (all columns) ([0.015] each)
  - Common language: 1.241***, 1.241***, 1.241***, 1.272*** ([0.034] each)
  - Colonial history: 0.849***, 0.850***, 0.850***, 0.832*** ([0.086], [0.086], [0.086], [0.084])
  - Geographical contiguity: 1.210***, 1.207***, 1.208***, 1.163*** ([0.067], [0.067], [0.067], [0.066])
  - Population, origin: 0.523***, 0.517***, 0.514***, 0.498*** ([0.022], [0.023], [0.023], [0.022])
  - Population, destination: 0.526***, 0.562***, 0.514***, 0.576*** ([0.023], [0.024], [0.024], [0.022])
  - REER, destination: 0.001*** (all columns) ([0.000] each)
  - Life expectancy, destination: 0.097, 0.039, 0.214, 0.209 ([0.070], [0.078], [0.081], [0.068])
- Disease variables (where included):
  - Ebola, Origin: 0.065*** ([0.005])
  - Ebola, Destination: 0.089*** ([0.008])
  - SARS, Origin: 0.387*** ([0.092])
  - SARS, Destination: 0.078*** ([0.104])
  - Malaria and Yellow fever coefficients reported as small/insignificant in these specifications.
- Sample sizes and fit:
  - Number of observations: 210,221; 210,221; 210,221; 213,645 (by column)
  - Origin FE: Yes; Destination FE: Yes; Year FE: Yes
  - Pseudo R2: 0.83, 0.83, 0.83, 0.84

### Robustness checks and sub-samples (2SLS-IV)
- Truncated sample, sub-sample (1995–2007), and additional controls shown.
- Selected coefficients (three reported columns):
  - Real GDP, origin: 0.961***, 1.259***, 1.023*** ([0.013], [0.028], [0.018])
  - Real GDP, destination: 0.808***, 0.751***, 0.967*** ([0.015], [0.028], [0.019])
  - Distance: 1.632***, 1.688***, 1.704** ([0.015], [0.021], [0.018])
  - Common language: 1.202***, 1.170***, 1.233*** ([0.033], [0.046], [0.039])
  - Geographical contiguity: 1.116***, 1.232***, 1.261** ([0.072], [0.084], [0.074])
  - Population, origin: 0.478***, 1.224***, 0.554*** ([0.022], [0.057], [0.033])
  - Population, destination: 0.478**, 0.305***, 0.625*** ([0.023], [0.057], [0.032])
  - REER, destination: 0.001**, 0.001**, 0.001** ([0.000] each)
  - Life expectancy, destination: 0.151, 0.060, 0.711*** ([0.078], [0.136], [0.194])
- SARS coefficients in robustness checks:
  - SARS, Origin: 0.224***, 0.281***, 0.224*** ([0.085], [0.041], [0.095])
  - SARS, Destination: 0.028***, 0.016***, 0.011*** ([0.091], [0.050], [0.103])
- Number of observations: 176,489; 96,416; 111,591
- Adjusted R2: 0.79, 0.83, 0.83

### Estimations by income group and region (2SLS-IV)
- Columns: Advanced; Developing; Africa; Asia; Europe; Latin America; Middle East.
- Selected coefficients (showing heterogeneity across regions):
  - Real GDP, origin: 1.008***; 0.935***; 0.860***; 1.044***; 0.754***; 0.889***; 1.169*** ([0.017], [0.020], [0.044], [0.054], [0.044], [0.035], [0.052])
  - Real GDP, destination: 1.083***; 0.744***; 0.304***; 1.201***; 0.575***; 0.744***; 0.371*** ([0.044], [0.024], [0.040], [0.077], [0.045], [0.047], [0.060])
  - Distance: 1.351***; 1.821***; 1.512**; 1.932***; 1.518***; 1.737***; 1.482*** ([0.026], [0.019], [0.053], [0.077], [0.085], [0.051], [0.072])
  - Common language: 0.584***; 1.404***; 1.170***; 0.697***; 0.863; 1.451***; 0.781*** ([0.054], [0.040], [0.058], [0.112], [0.406], [0.064], [0.126])
  - Geographical contiguity: 0.480***; 1.307***; 1.126**; 0.936***; 1.525***; 0.970***; 1.662*** ([0.106], [0.079], [0.148], [0.185], [0.167], [0.150], [0.198])
  - Population, origin and destination show large regional variation (see table values).
  - REER, destination: 0.006*** (Advanced); 0.000*** (Developing); 0.000 (Africa); 0.002*** (Asia); 0.000 (Europe); 0.000*** (Latin America); 0.000 (Middle East) ([0.000]–[0.001] as listed)
  - Life expectancy, destination: 5.847*** (Advanced), 0.585*** (Developing), 0.366* (Africa), 1.668** (Asia), 3.911*** (Europe), 0.745 (Latin America), 0.887 (Middle East) ([0.528], [0.087], [0.146], [0.554], [1.264], [0.420], [0.659])
- SARS coefficients by region include negative estimates with varying significance (see table for exact values).
- Number of observations by region: 70,721; 139,500; 36,232; 23,922; 23,794; 33,750; 21,802
- Adjusted R2 by region: 0.88; 0.85; 0.81; 0.86; 0.85; 0.85; 0.81

### Out-of-sample forecast performance
- Training period: 1995–2014. Test period: 2015–2017.
- PPML models (Standard, Ebola, SARS):
  - MAE: 5.103; 5.086; 4.954
  - RMSE: 5.806; 5.791; 5.675
  - U-Theil: 0.625; 0.624; 0.613
- 2SLS-IV models (Standard, Ebola, SARS):
  - MAE: 1.168; 1.165; 1.096
  - RMSE: 1.675; 1.652; 1.561
  - U-Theil: 0.118; 0.116; 0.111
- Interpretation: The model with the lowest MAE, RMSE, and U-Theil is considered to better forecast accuracy.

### Summary of new features in the theoretical model (GIMF with tourism)
- GIMF: annual, multi-region, micro-founded general equilibrium model of the global economy (documentation referenced elsewhere).
- Tourism sector treatment:
  - Tourism bundle produced like consumption and investment goods from tradable and nontradable goods, with an added productivity term capturing pandemic costs (health costs, social distancing, new technologies).
  - Tourism bundle can be consumed domestically (domestic households) or by foreign visitors (tourism exports/imports tracked bilaterally).
  - Households have a two-item consumption bundle: tourism services and other goods and services.
  - Tourism has a price and affects the consumer price index (CPI) inflation.
  - Trade in tourism services is tracked bilaterally between all regions.
- Additional trade detail:
  - Consumption and investment goods are always treated separately (unlike standard GIMF).
  - Non-tariff barriers (NTBs) present where country A imposes NTB but country B bears the cost in production and passes it to importing consumers via higher prices (contrast with tariffs).

### Calibration of the Asia-Pacific and Western Hemisphere GIMF variants
- Two structurally identical eight-region versions used: one focused on Asia-Pacific; the other on the Western Hemisphere, with overlapping global regions.
- Regions and their compositions are enumerated in the source (region definitions provided).
- Calibration draws on 2018-consistent stylized data and assumptions on long-term stock ratios.
- Tourism data (exports, imports, consumption, production) used to calibrate tourism production shares; tourism exports and production shares are largest as a share of GDP in the Pacific Islands and the Caribbean.

### Key national accounts ratios in GIMF (selected table entries)
- Share of Global GDP (%, US$): Caribbean 0.15; Pacific Islands 0.01; ASEAN-5 3.04; Central America 0.31; Australia/New Zealand 1.91; Canada 2.01; China 15.74; Europe 21.79; Latin America 5.26; United States 24.23
- Domestic demand: Household Consumption (% of GDP): Caribbean 63.9; Pacific Islands 51.1; ASEAN-5 61.0; Central America 65.3; Australia/New Zealand 57.4; Canada 56.8; China 47.5; Europe 57.0; Latin America 65.6; United States 65.2
- Tourism consumed (% of GDP): Caribbean 5.0; Pacific Islands 8.0; ASEAN-5 6.0; Central America 4.0; Australia/New Zealand 5.0; Canada 5.0; China 3.6; Europe 5.5; Latin America 2.6; United States 2.0
- Tourism produced (% of GDP): Caribbean 18.3; Pacific Islands 28.9; ASEAN-5 7.8; Central America 9.6; Australia/New Zealand 5.3; Canada 3.9; China 2.0; Europe 7.3; Latin America 1.9; United States 2.8
- Trade and tourism service shares for goods exports/imports, consumption, investment, intermediate, and tourism service exports/imports are reported in the table (see source for full matrix).

### Elasticities and adjustment parameters
- Tourism-related elasticities (Annex Table 3.2):
  - Elasticity between Consumption and Tourism: 0.9
  - Elasticity Domestic/Foreign Tourism: 0.9
  - Elasticity Foreign Tourist Destinations: 1.5
- Goods trade elasticities (between domestic/imported goods and different regions’ goods):
  - Consumption: 1.5 (Domestic/Imported Goods), 1.5 (Different Regions’ Goods)
  - Investment: 1.5, 1.5
  - Intermediates: 1.5, 1.5
- Real rigidities:
  - Set to 1 across all regions for most variables; consumption by saving households set to 2.
  - For COVID-19 pandemic shocks, real rigidities in the tourism sector set to zero to account for unusually rapid adjustment.
- Nominal adjustment costs:
  - Set the same across domestic prices but higher in Europe.
  - Vary widely across regions for imports; large import markets (United States) have high adjustment costs; smaller markets (Caribbean, Pacific Islands) are price takers with low adjustment costs.
  - For COVID-19 shocks, nominal adjustment costs for domestic and imported tourism prices greatly reduced in all regions.
- Calibration of nominal adjustment costs (Annex Table 3.3) — baseline and COVID-19 shock values (selected entries):
  - Real wage, Baseline/COVID-19 Shock: United States 100/100; Europe 150/150; China 100/100; Canada, Other Asia-Pacific, Latin America 100/100
  - Tourism price, Baseline/COVID-19 Shock: United States 100/50; Europe 150/50; China 100/50; Canada, Other Asia-Pacific, Latin America 100/50
  - Price of imports of Goods, Baseline/COVID-19 Shock: United States 100/100; Europe 50/50; China 25/25; Canada, Other Asia-Pacific, Latin America 10/10
  - Price of imports of Tourism, Baseline/COVID-19 Shock: United States 100/8/50? (table entries indicate large reductions; see source table for exact formatting and values)

### Summary of assumptions underpinning GIMF scenario use
- 1. All agents in the model have perfect foresight.
- 2. All regions share the same economic structures, differing only through parameterization and calibration.
- 3. Model is annual frequency; some short-run dynamics (particularly first-year investment) are less detailed.
- 4. Baseline calibration based on parameter values consistent with 2018 for capital stock, government debt and deficit, net foreign assets, current account balance, national accounts aggregates, and trade flows and services data.
- 5. Model includes non-linearities in the financial accelerator and potential non-linearities in monetary policy (e.g., encountering the zero-interest-rate floor), otherwise approximately linear for small shocks.

*Source: Annex 2. Applying A Gravity Model to Predict Post-Pandemic Tourism Flows (text and tables as provided).*

### 6. The real exchange rate is a “jumper,” adjusting immediately in the first

### 6. The real exchange rate is a “jumper,” adjusting immediately in the first year to shocks

### Key model assumptions and market structure
- The real exchange rate adjusts immediately in the first year to shocks, following the forward-looking, risk-adjusted uncovered interest rate parity condition that equates the forward sum of national-US interest rate differentials with the one-year-forward difference in the nominal exchange rate.
- No financial friction is included in the exchange rate equation to force net foreign asset positions to a steady state; net foreign asset positions and their dynamics solve endogenously within the OLG framework.
- GIMF contains only a financial accelerator (full general equilibrium form with non-linearities) and assumes complete domestic ownership of firms.
- All net foreign asset positions are denominated in US dollars, in all countries.

### Monetary and fiscal policy assumptions
- Monetary Policy:
  - Monetary policy is passive and relies only on its inflation targeting rules (interest rate reaction functions).
  - No additional monetary policy stimulus is applied to offset shocks to tourism.
- Fiscal Policy:
  - Government follows through on pre-COVID-19 spending plans for government consumption, infrastructure investment and household transfers; automatic stabilizers are active and revenues move with the state of the economy.
  - Some fiscal features are altered when modelling fiscal stimulus shocks in Box 1.
  - The deficit will vary over time: growing during recessionary phases and shrinking during expansions due to automatic stabilizers, variability in revenues, and unchanged spending plans.
  - For model stability, after 20 years the government pursues its pre-COVID-19 debt-and deficit-to-GDP targets. The deficit-to-GDP target is achieved in the short term by cutting transfers to households; combined with ongoing economic growth, the debt-to-GDP ratio returns to its target in the longer term.

### Benchmark scenario (technical implementation)
- Scenario composition:
  - Combines shocks to consumer preferences for all tourism and domestic versus foreign tourism.
  - Adds a permanent negative productivity shock in all regions’ tourism sectors.
  - Simulation starts in Year 1 using the existing steady state (representative of the pre-COVID-19 global economy) as control.
  - Tourism collapses in Year 1 and then recovers, subject to permanent economic scarring.
- Technical description:
  1. Tourism is reduced to close to zero in Year 1 by:
     - Reducing the bias toward other consumption over tourism (ALPHA_C) to almost zero within each region.
     - Reducing the bias toward domestic tourism over foreign tourism (ALPHA_SH) to almost zero within each region.
  2. Tourism reopens from Year 2 to Year 14, with emphasis on domestic tourism:
     - Return ALPHA_C to original values with a decay rate of 0.65.
     - Return ALPHA_SH to original values with a decay rate of 0.65.
  3. Economic scarring:
     - Add a permanent −10 percent shock starting in Year 1 to the level of sectoral productivity in the production function for tourism (AA_S) in each region.
- Measurement note:
  - Even without productivity shock, real GDP may not return to its original level because of shifts in relative prices and component inflation rates versus pre-COVID-19 forecasts; the relevant model-based measure is real income, which does return to its pre-COVID-19 forecasted level in the long term.

### Downside scenario (technical implementation)
- Built on top of the benchmark; assumes more severe outcomes beginning in Year 2 (unanticipated at Year 1).
- Decomposed into three layers: additional scarring, slower recovery of preferences, tourists differentiating by destination (NTBs).
- Technical description:
  1. More economic scarring:
     - As of Year 2, sectoral productivity in tourism (AA_S) is at −20 percent (−10 percent in the benchmark).
  2. Slower recovery of tourist preferences:
     - Starting in Year 2, reduce speed of recovery for ALPHA_C using a decay rate of 0.80 instead of 0.65.
     - Starting in Year 2, reduce speed of recovery for ALPHA_SH using a decay rate of 0.80 instead of 0.65.
  3. Tourists differentiate among destinations based on healthcare quality via NTBs added to the tourism sector (NTB_S_[visiting region]_[visited region]), starting in Year 2. Shocks placed against visiting regions as follows:
     - 5 percent for ANZ, CAN, EUR, USA
     - 7.5 percent for OAP
     - 10 percent for ASE, CHN, CRB, LAM, RC1, RC2
     - 15 percent for CAM, OWH, PIC

### Upside scenario (technical implementation)
- Built on top of the benchmark; assumes more favorable outcomes even from Year 1.
- Key features:
  1. No economic scarring:
     - No change in level of sectoral productivity in tourism (benchmark is −10 percent).
     - Real income will return to original value; real GDP may still deviate long term due to relative price shifts.
  2. Faster recovery of tourist preferences:
     - Starting in Year 2, higher speed of recovery for ALPHA_C using decay rate of 0.20 instead of 0.65, then pause in Year 3 by using decay rate of 1.00.
     - Starting in Year 2, higher speed of recovery for ALPHA_SH using decay rate of 0.20 instead of 0.65, then pause in Year 3 by using decay rate of 1.00.
     - Starting in Year 4, increase speed of recovery for ALPHA_C with decay rate of 0.40 instead of 0.65.
     - Starting in Year 4, increase speed of recovery for ALPHA_SH with decay rate of 0.40 instead of 0.65.

### Differing extents of economic scarring (technical implementation)
- Four options for scarring represented by productivity shocks (AA_S) starting in Year 1:
  - +10 percent
  - 0 percent
  - −10 percent (benchmark)
  - −20 percent (alternative)
- Outcomes:
  - Range of real GDP outcomes is widest for heavily tourism-dependent regions (Pacific Islands and Caribbean), up to more than 13 percentage points after 10 years; much less for others (about 1 percentage point for China).

### Differing speeds of recovery (technical implementation)
- Two simulations starting in Year 1 using benchmark as control; after five years preferences for tourism are back at either 67 percent or 99 percent of pre-COVID-19 levels, versus benchmark level of 88 percent.
- Technical description:
  1. Starting in Year 2, reduce (increase) speed of recovery for ALPHA_C using decay rate of 0.80 (0.40) instead of 0.65.
  2. Starting in Year 2, reduce (increase) speed of recovery for ALPHA_SH using decay rate of 0.80 (0.40) instead of 0.65.

### Tourists differentiating by destination (technical implementation)
- Tourists permanently differentiate among destinations by healthcare quality, represented by NTBs on tourism beginning in Year 2.
- Technical description and NTB shocks placed against visiting regions as follows:
  - 5 percent for ANZ, CAN, EUR, USA
  - 7.5 percent for OAP
  - 10 percent for ASE, CHN, CRB, LAM, RC1, RC2
  - 15 percent for CAM, OWH, PIC

*Source: IMF staff calculations from Annex 3, "Analyzing Macroeconomic Scenarios Using GIMF," Tourism in the Post-Pandemic World (chapter content).*

### 2. Same as the first scenario, but those regions facing 10 percent NTBs have

### tfotppwea - 2. Same as the first scenario, but those regions facing 10 percent NTBs have

### Policy Scenario 1: Short-Term Fiscal Stimulus — design and timing
- Fiscal stimulus packages run for three years.
- Fiscal stimulus size: 3 percent of GDP for the first two years, and then halved in the third year (implying 1.5 percent of GDP in Year 3 for measures that are halved).
- After Year 3, fiscal instruments return to former settings except government investment, which may be sustained to maintain a new higher infrastructure capital stock.
- Two alternative fiscal packages are simulated for the ASEAN-5 (ASE) in the Asia-Pacific version of GIMF: a lower-multiplier package and a higher-multiplier package.

### Technical description — Lower-multiplier fiscal package (shocks introduced in Years 1–3)
- Year 1 and Year 2 shocks:
  - +1.0 percent of GDP shock on government consumption (GOVCONS) through E_GOVCONS (scaled by GDP and GOVCONS from the benchmark scenario).
  - +1.5 percent of GDP shock on general lump-sum transfers (TRANSFERS).
  - −0.5 percent of GDP shock on labor income taxes for saving households (TAU_L_OLG) through TAULOLGBAR (scaled by GDP and TAXBASE_L_OLG from the benchmark scenario).
  - Increase deficit target to GDP ratio (GOVSURSTAR) through E_GOVSURSTAR by 3 (which is 3 percent of GDP).
- Year 3: each shock cut in half:
  - GOVCONS shock: +0.5 percent of GDP.
  - TRANSFERS shock: +0.75 percent of GDP.
  - TAU_L_OLG shock: −0.25 percent of GDP.
  - GOVSURSTAR increase cut to 1.5 percent of GDP.

### Technical description — Higher-multiplier fiscal package (shocks introduced in Years 1–3)
- Year 1 and Year 2 shocks:
  - +1.5 percent of GDP shock on government infrastructure investment (GOVINV) through E_GOVINV (scaled by GDP and GOVINV from the benchmark scenario).
  - +1.0 percent of GDP shock on targeted lump-sum transfers (TRANSFERS_TARG_RAT) and set TRANSFER-SHARE=1 so transfers are directed only to liquidity-constrained households.
  - −0.5 percent of GDP shock on consumption value-added tax (TAU_C) through TAUCOLGBAR (scaled by GDP and TAX-BASE_C from the benchmark scenario), which will automatically flow through to TAUCLIQSTAR.
  - Increase deficit target to GDP ratio (GOVSURSTAR) through E_GOVSURSTAR by 3 (which is 3 percent of GDP).
- Year 3: each shock cut in half:
  - GOVINV shock: +0.75 percent of GDP.
  - TRANSFERS_TARG_RAT shock: +0.5 percent of GDP.
  - TAU_C shock: −0.25 percent of GDP.
  - GOVSURSTAR increase cut to 1.5 percent of GDP.

### Key policy-design implications (from the scenario construction)
- Mixing government consumption, transfers (general or targeted), tax cuts, and government investment enables alternative multiplier outcomes.
- Targeting transfers to liquidity-constrained households (TRANSFER-SHARE=1) is part of the higher-multiplier strategy.
- Infrastructure investment is modeled as a persistent capital stock increase beyond the three-year shock horizon.

---

### Policy Scenario 2: Tourism Bubble — structure and variants
- Two-simulation approach:
  1. A two-year variant of the benchmark scenario where Year 1 tourism shocks extend into Year 2 before decaying.
  2. A travel bubble between Australia and New Zealand (ANZ) and the Pacific Islands (PIC), implemented on top of the first simulation.

### Technical description — First simulation (severity enhancement)
- Begin in Year 2 using the benchmark scenario as control.
- Keep tourism close to zero in Year 2 across all regions by:
  - Maintaining bias toward other consumption over tourism (ALPHA_C) at Year 1 value.
  - Maintaining bias toward domestic tourism over foreign tourism (ALPHA_SH) at Year 1 value.
- Tourism reopens from Year 3 to Year 14 with greater emphasis on domestic tourism:
  - Return ALPHA_C to original values with a decay rate of 0.65.
  - Return ALPHA_SH to original values with a decay rate of 0.65.

### Technical description — Second simulation (ANZ–PIC travel bubble starting Year 2)
- Control case: the first simulation above.
- Year 2 adjustments specific to ANZ and PIC:
  - Remove the shock from ALPHA_C in ANZ and PIC to allow full resumption of tourism as a share of consumption in those two regions.
  - Return ALPHA_SH to original values with decay rates: 0.80 in ANZ and 0.65 in PIC.
  - For Year 2 only, add NTBs on imports of tourism services by ANZ from other regions (NTB_S_ANZ_[visited region]) of 200 percent.
  - For Year 2 only, add NTBs on imports of tourism services by PIC from other regions (NTB_S_PIC_[visited region]) of 200 percent.

### Key scenario parameters to note (numeric fidelity preserved)
- Decay roots used in benchmark and variants: 0.65 (benchmark), alternative roots in related scenarios: 0.8, 0.9, 0.80 (as specified elsewhere for regions facing 10 percent and 15 percent NTBs).
- NTB shocks in bubble: 200 percent (Year 2, ANZ and PIC).

---

### Non-tourism exports in tourism-dependent economies — observations and examples
- Several tourism-dependent countries have developed other export sectors, helping mitigate tourism shocks.
- Examples and comparative observations:
  - Some countries with abundant natural resources (Chile and Mexico cited) have fuel/mineral commodity exports per capita in the world’s top quartile.
  - Some countries have significant non-commodity goods exports per capita (Thailand, Panama, Chile, and Mexico cited among the world’s upper quartile).
  - Several small states (population below 1 million) have notable manufacturing exports (examples: Bahamas, Barbados, Saint Kitts and Nevis, and Saint Lucia exporting medicaments, medical goods and equipment, ships and vessels, and electronics).
  - Natural resource–based exports (fisheries, agriculture) are also a diversification option for some small states.
  - Some countries have developed other service exports per capita; many small-population countries are in the world’s upper half for non-tourism services per capita, with Central America and the Caribbean containing many in the top quartile.

- Caveats:
  - Non-tourism services exports can be concentrated in financial services that benefit from very low taxation and/or passport sales; these face uncertainty due to international regulatory changes and volatile demand for second-citizenship programs.
  - Medical and education tourism services are suggested as more stable diversification options.

---

### Export development potential, complexity, and quality upgrade — findings
- Export diversification and complexity are positively associated with income per capita at least until advanced-economy levels.
- Economic Complexity Index (ECI) captures diversity and ubiquity of a country’s export basket; several tourism-dependent countries have relatively low ECI, constraining options for new-product development.
- Complexity Outlook Index (COI) measures how many complex products are near a country’s current capabilities; COI identifies diversification potential:
  - Countries with few nearby opportunities (require stronger policies): Cambodia, Dominican Republic, Honduras, Jamaica, Mongolia, Nicaragua, Panama (noting the source text lists these as examples requiring stronger policies).
  - Thailand is highlighted as having many nearby diversification options due to existing production capacity and know-how.
- Product-space analysis (examples: Costa Rica, Jamaica, Thailand):
  - Identifies related products with potential (industrial machinery, electrical machinery and equipment) based on distance and product complexity.
  - Bubble-size in the analysis reflects global market size for each good.

- Quality-upgrade opportunities:
  - Quality upgrade of existing products can be a feasible diversification route where moving to new sectors is costly.
  - Quality index (per Henn, Papageorgiou, and Spatafora (2013)) measures unit values of exported goods adjusted for production-cost differences and selection bias.
  - Several tourism-dependent economies have significant scope for quality upgrade; example: Cambodia has large potential for quality upgrades in clothing and footwear and machinery products.
  - Small states and specific sector positions are highlighted in the sectoral quality ladders (countries and sectoral shares illustrated in the source figures).

---

*Source: IMF staff calculations (from the chapter/annex material provided in the source PDF).*

### Annex Figure 4.4. Quality Ladders

### Annex Figure 4.4. Quality Ladders

### Key findings from quality ladders (exports by sector)
- Among Latin American and Caribbean countries, Argentina, Chile, and Honduras show potential to upgrade quality across important export sectors, such as clothing, food, and manufacturing goods.
- Micro-states, such as St. Vincent and the Grenadines, and Saint Kitts and Nevis, produce machinery and manufactured goods that are at the world frontier of quality, including ships and boats and measuring instruments.
- The figure presents Export Quality Index values across sectors (Food and live animals; Machinery; Crude materials; Chemicals; Manufacturing goods; Misc. manufacturing articles) for:
  - Asia-Pacific (Small States)
  - Asia-Pacific (Other)
  - Latin America (Small States)
  - Latin America (Other)

### Determinants of a country’s ability to diversify and develop complex exports
- The literature identifies three main groups of determinants:
  - Distance/proximity to other economies.
  - Productivity-related variables.
  - Unit labor costs.
- Proximity insights:
  - A proximity index is constructed as the sum of the size of partner economies weighted by the inverse of their distance.
  - For Latin America and the Caribbean, only Argentina, Chile, and Uruguay are relatively remote; Caribbean countries and Mexico (the most tourism-dependent in the region) have an index above the world’s median, giving greater scope to integrate into value chains.
  - In Asia and the Pacific, many countries have an index of proximity to markets below the world’s median, driven by highly tourism-dependent, small, and remote Pacific islands.
- Productivity- and institution-related variables:
  - Governance (World Bank Worldwide Governance Indicator) is relatively sound among most tourism-dependent countries but positioned just above the global median, indicating scope for enhancement.
  - Educational quality (UN Human Development Index education sub-index) is one of the most significant and robust determinants of export diversification and complexity; many tourism-dependent countries—especially in Asia-Pacific—have educational attainment below the world median.
  - Infrastructure data coverage is limited for several small states; for covered countries the Global Competitiveness Index’s infrastructure pillar indicates relatively weak infrastructure quality (in the world’s lowest quartile) for some tourism-dependent countries.
  - Many tourism-dependent countries have relatively high import tariffs, averaging about and above 10 percent. Only Central American countries as well as Chile have average tariffs among the world’s lowest quartile.
  - Labor market efficiency indicators show room for improvement; several Latin American countries rank well below the world’s median on the Global Competitiveness Report labor market efficiency index, often due to significant labor market rigidities. Many small Caribbean countries have strong unionization and high collective bargaining coverage.

### Policy and institutional choices and recommendations
- Leverage proximity and connectedness:
  - Use proximity to large markets (for example, the United States or China) to integrate into global value chains where feasible.
  - Effectively shorten distance to other economies by enhancing connectedness at all levels: reduce trade policy barriers, enhance trade facilitation, strengthen transport infrastructure, invest in top-notch communication technology (particularly internet connectivity), and foster technological diffusion.
- Strengthen productivity-related policies:
  - For remote and small tourism-dependent economies, focus policies to raise productivity—potentially with emphasis on services sectors, which are less dependent on distance to other markets.
  - Develop sector-specific infrastructure to support promising non-tourism exports and invest in natural disaster–resilient infrastructure to protect tourism and non-tourism activities.
  - Given fiscal constraints, consider private sector participation for infrastructure development.
- Improve governance and institutional quality:
  - Enhance government effectiveness and combat corruption—both are significantly associated with export diversification.
  - Address crime indicators (notably in some Caribbean and Central American countries) that limit export development, including tourism.
- Invest in human capital and labor market policies:
  - Improve educational quality and provide retraining to support labor reallocation in scenarios of declining tourism.
  - Reform labor market rigidities to improve labor market efficiency, facilitate sectoral reallocation, and avoid raising effective unit labor costs.
  - Design public sector labor policies to target efficiency by linking wages and employment to productivity to reduce incentives for outward migration of better-trained workers.
- Trade and tax policy adjustments:
  - Reduce bias in favor of the tourism sector (for example, widespread tax exemptions for tourism imports) and lower trade barriers across the board to support non-tourism exporters.
  - Ensure taxation does not penalize non-tourism sectors.
  - Promote trade openness, including intraregional trade, which is especially critical for small and microstates to offset lack of economies of scale.

*Source: Annex Figure 4.4 and accompanying text, "Annex 4. Assessing Export Development Potential in Tourism-Dependent Economies," from the provided PDF content.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2021/english/tfotppwea.pdf_