## apd-reo-background-papers-1018 ## Source details **Canonical URL:** [apd-reo-background-papers-1018](https://www.imf.org/-/media/files/publications/reo/apd/2018/apd-reo-background-papers-1018.pdf) ## Other formats - [Markdown version](/-/media/files/publications/reo/apd/2018/apd-reo-background-papers-1018.pdf.md) - [Structured JSON version](/-/media/files/publications/reo/apd/2018/apd-reo-background-papers-1018.pdf.json) --- ### Global outlook and risks - World growth projected to decelerate to 2.1 percent in 2019 as output gaps close and monetary policy becomes less accommodative. - Emerging market and developing economies projected to grow at 4.7 percent in 2018–19. - Risks tilted more to the downside near term relative to April 2018 WEO; contributing factors include: - Weak growth outturns in the first half of 2018 in several large economies. - Moderation in high-frequency economic indicators. - Tighter financial conditions in vulnerable economies. - Risk of further escalation of trade actions and geopolitical tensions. - Delays in addressing inequality and climate change. ### Regional outlook: Asia — summary and headline projections - Asia: real GDP growth projected at 5.6 percent in 2018 (unchanged from April) and 5.4 percent in 2019 (down by 0.2 percent from April). - Inflation (Asia): 2.7 percent in 2018 and 2.9 percent in 2019. - Current account (Asia): 1.5 percent of GDP in 2018 and 1.4 percent of GDP in 2019. ### Regional developments and selected country highlights - Asia-wide: - GDP growth outturns H1 2018 softer than in 2017 for advanced economies; robust in emerging economies. - Headline inflation inched up in several economies due to higher oil prices and currency depreciation. - Current account surpluses narrowed; many economies decreased foreign reserves in 2018. - China: - Investment growth slowed from 6 percent year-over-year in June to 4.1 percent in August. - Unemployment rose from 4.8 percent in June to 5 percent in August. - Headline CPI inflation around 2 percent. - Credit growth slowed to 10.1 percent year-over-year in August (down from around 14 percent at end-2017). - Renminbi depreciated 7 percent against the US dollar since mid-June. - Equity prices declined substantially; money market rates declined reflecting monetary easing. - Japan: - Temporary contraction in Q1 2018; rebound in Q2 2018. - Headline inflation 1.3 percent year-over-year in August; core inflation 0.4 percent year-over-year in August. - India: - GDP growth 6.7 percent in FY2017/18; 7.7 percent year-over-year in Q1 2018. - Headline inflation averaged 3.6 percent in FY2017/18 (a 17-year low). - Inflation 3.7 percent year-over-year in August (down from 4.2 percent in July). - Korea: - Growth 3.1 percent in 2017 and 2.9 percent in H1 2018. - Headline and core inflation 1.4 and 0.9 percent year-over-year respectively in August. - Current account surplus 5.1 percent of GDP in 2017; narrowed further in H1 2018 by around 2.5 percentage points of GDP (year-over-year). - Australia: - Annual growth 3.3 percent in H1 2018; consumer prices rose 2.1 percent year-over-year in Q2 2018; core inflation 1.9 percent year-over-year. - New Zealand: - GDP growth 2.8 percent year-over-year in Q2 2018; headline inflation 1.5 percent; core inflation 1 percent. - Hong Kong SAR: - Strong GDP growth in H1 2018; exports expanded 8 percent year-over-year in H1 2018 with a 5 percent increase in volumes. ### ASEAN and Southeast Asia: selected statistics - Indonesia: - Growth 5.2 percent year-over-year in H1 2018 (from 5.1 percent in 2017). - Rupiah depreciated by 10 percent since January; equity prices fell by 8 percent. - 10-year government bond yields rose by 180 basis points. - Bank Indonesia policy rate raised to 5.75 percent. - Thailand: - GDP increased 4.8 percent year-over-year in H1 2018. - Headline inflation 1.6 percent year-over-year in August; core inflation 0.8 percent. - Singapore: - Real GDP grew 3.6 percent in 2017 and 4 percent year-over-year in H1 2018. - Headline inflation 0.7 percent year-over-year in August. - Malaysia: - Real GDP grew 4.9 percent year-over-year in H1 2018 (vs 5.9 percent in 2017). - Headline inflation 1.6 percent year-over-year in H1 2018 (down from 3.8 percent in 2017). - Current account surplus 2.7 percent of GDP in H1 2018 (from 3 percent of GDP in 2017). - Philippines: - Growth 6.7 percent in 2017 and 6.3 percent year-over-year in H1 2018. - Inflation rose to 6.4 percent year-over-year in August (exceeding 2 to 4 percent target band). - Vietnam: - Growth 7.1 percent year-over-year in H1 2018. - Inflation 4 percent year-over-year in August; core inflation 1.5 percent year-over-year. - Myanmar, Cambodia, Lao P.D.R., Brunei: - Myanmar: rebound in 2017/18 toward previous potential. - Cambodia: growth around 7 percent; inflation 3 percent year-over-year in June. - Lao P.D.R.: moderation; inflation expected to remain at around 2 percent in 2018. - Brunei Darussalam: real GDP growth 1.3 percent in 2017; inflation -0.2 percent in 2017, turned positive in early 2018. ### Other frontier and Pacific economies - Bangladesh: - Real GDP growth in FY2018 projected to have accelerated to 7.5 percent. - Inflation eased to 5.5 percent by end of FY2018 after flood-related rises. - Nepal: - Growth 6.3 percent in FY2017/18; inflation fell to 4.2 percent. - Sri Lanka: - Gradual recovery under way after subdued growth in 2017. - Mongolia: - Growth accelerated from 5.1 percent in 2017 to 6.3 percent year-over-year in Q1 2018. - Fiscal: 3.4 percent of GDP primary surplus in Q1 2018. - Pacific Island countries: - Growth around 2.6 percent in 2017. - Fiji: growth jumped to 3 percent (recovery from Cyclone Winston). - Papua New Guinea and Timor-Leste: subdued growth due to weak commodity prices. - Cyclone Gita caused widespread destruction in Tonga and Samoa. ### Financial conditions and capital flows in Asia - Financial conditions tightened in some large Asian economies due to US policy normalization, rising global trade tensions, and volatility in China and other large EMs. - Cumulative portfolio inflows in 2018 far below 2016–17 levels because of large outflows during Q2 2018. - Nonresident sales of portfolio debt securities increased as the US dollar appreciated. - Asian equity indices and exchange rates negatively affected by trade tensions; bond yields and spreads generally increased. - Policy responses: some central banks raised policy rates; others intervened to support domestic currencies. - Recent volatility comparable in some respects to the taper tantrum; Asia affected much less than other regions and is more tied to developments in China. ### Trade tensions: modeled scenarios and macro impacts - Three tariff scenarios modeled (baseline, escalation, automobile sector) with additional confidence and financial effects layers. - Baseline elements include: - 25 percent tariff on steel and 10 percent tariff on aluminum by the United States; 25 percent US tariff on $50 billion of China imports; announcement on September 17 of further $200 billion (10 percent rising to 25 percent by year-end) on China; assumed Chinese retaliation on $60 billion. - Escalation scenario: 25 percent tariff on all US imports from China (worth over $500 billion) and Chinese retaliation of 25 percent on all US imports. - Automobile sector scenario: adds 25 percent US tariff on cars, trucks, and parts; example impact treated as 3.75 percent tariff on all imported final goods if autos represent 15 percent of US imports of final goods. - Timing: 2018 effects cover six months for baseline tariffs (or three months for some measures); escalation and auto scenarios assumed from 2019. - Modeled peak losses if all channels (tariffs plus confidence and financial effects) are in play: - Peak GDP losses of 1.6 percent in China. - Peak GDP losses close to 1 percent in the United States. - Peak GDP loss for Asia as a whole would be 0.9 percent. - Under the full scenario: - China: real GDP falls by 1.6 percent; investment hit by 4.6 percent; consumption falls by 1.9 percent. - Korea: real GDP could fall by around 0.9 percent. - India, Indonesia, Japan: real GDP could fall by around 0.7 percent. - Philippines: around 0.6 percent; Australia, Malaysia, Thailand: around 0.5 percent; New Zealand: 0.3 percent. - Trade volumes weaker in Asia: Japan around 2.2 percent weaker; China more than 1.5 percent weaker; Korea almost 1.2 percent weaker. ### Channels of trade-tension impacts - Three channels modeled: 1. Trade channel: higher tariffs depress relative demand for imports and reduce GVC participation. 2. Confidence channel: depressed business confidence lowers private investment in 2018–19. 3. Financial channel: higher corporate spreads peak in 2019 and further drag on private investment. - Broader effects include trade diversion, supply-chain contraction, sectoral job displacement (e.g., metals, electrical equipment, transport), and broad-based US dollar appreciation with renminbi depreciation. ### Trade costs, NTBs, and liberalization scenarios - Trade costs components highlighted: policy-related NTBs account for 60 to 90 percent of trade costs; tariffs and natural trade barriers account for the remaining 10 to 30 percent (Duval and Utoktham 2011). - Three liberalization scenarios evaluated with RTM and GIMF: 1. "China Opens": China eliminates goods tariffs and reduces services NTBs for all partners. 2. "Within Asia": All of Asia eliminates goods tariffs and reduces services NTBs for Asian partners. 3. "Asia Opens": All of Asia eliminates goods tariffs and reduces services NTBs for all partners; variant "Asia Opens + FDI" includes FDI liberalization. - Calibration: - Services NTBs reduced by 10, 20, or 30 percent depending on initial NTB categorization (Low, Average, High). - Goods tariffs reduced to lowest value among Asian economies in each sector (most sectors to zero). - Estimated long-term impacts (scenario 3 with FDI liberalization): - Exports of services rise by 17 percent (world trade). - Exports of goods rise by 9 percent (world trade). - Asia’s long-term GDP per capita rises by 7 percent under Asia opening without FDI liberalization and by 15 percent with FDI liberalization. - With FDI liberalization, China’s real GDP long-term gain: 9 percent total, of which six percentage points attributable to FDI liberalization. ### Policy implications on trade liberalization - Priorities: - Emphasize liberalization of services and easing of FDI restrictions to maximize gains. - Reduce trade costs beyond tariffs (customs, logistics, NTBs). - Use complementary domestic policies to smooth labor reallocation and manage transition costs (active labor market policies, social protection, education, targeted redistribution). - Negotiate reforms over years while deploying compensatory policies to share gains and overcome vested-interest resistance. ### Financial channels and US dollar funding costs - US monetary tightening put upward pressure on US dollar funding costs and the cross-currency basis. - Some economies exhibit persistent deviations from covered interest parity: - Consistently positive deviations: Australia, New Zealand, India. - Consistently negative deviations: Japan, Korea, Hong Kong SAR. - Small deviations: Singapore, Thailand, the Philippines. - Policy implications: - Central bank swap lines can mitigate dollar shortages. - Domestic capital market development for alternative funding instruments recommended. ### Corporate debt, vulnerabilities, and "zombie" firms - Emerging Asia corporate debt rose to about 114 percent of GDP in 2017:Q4 (from 71 percent in 2007:Q4). - Corporate vulnerability evidence: - Share of foreign currency debt declined over the past decade; average maturity of foreign currency debt decreased. - Altman Z-score mapping provided (verbatim ranges preserved in source). - Regression evidence: - A currency depreciation of 15 percent would reduce an implicit bond rating by one notch (example: A to A-). - "Zombie" firms: - Definition: 10 years old or older with interest coverage ratio < 1 for three consecutive years. - Japan: zombies ~10 percent of firms and account for 8 percent of total capital stock. - China: zombie capital share recently reached 6 percent. - Estimated benefits from reducing zombie shares: - Halving zombie capital share could increase annual TFP growth between 0.03 percent (Malaysia) and 0.09 percent (Thailand) in a representative four-digit industry. - Cutting zombie shares by half could increase tangible investment growth between 0.08 and 0.21 percentage points per year in representative-industry simulations. - Policy recommendations (financial sector and corporate): - Monitor vulnerable firms and systemically important corporates; fill data gaps on corporate finances. - Use macro- and microprudential policies to limit FX balance-sheet exposures and excessive leverage. - Prepare for an increase in corporate failures; reform insolvency regimes where needed. - Reign in excess credit growth (China) and support healthy bank balance sheets (India). ### Firm dynamism, productivity, and intangible investment - Decline in firm dynamism: share of young firms in Orbis sample fell from 45 percent in 2003 to 25 percent in 2015. - Young firms typically: - Lower initial TFP but higher TFP growth than older firms (except China where age effect positive but not statistically significant). - Smaller, faster-growing, lower leverage, invest more in intangible capital. - Intangibility and TFP: - Intangibility t-1 coefficients positive and significant in many regressions (preserved exact table estimates provided in source; e.g., Intangibility t-1 = 0.015*** (0.003) in pooled regression). - Subsamples show stronger intangibility effects where min(intangibility) > 1% or > 10%. - Regression highlights (preserved exact coefficients): - TFP t-1: −0.621*** (0.003) - Leverage t-1: 0.062*** (0.003) - Leverage^2 t-1: −0.007** (0.003) - Intangibility t-1: 0.015*** (0.003) - Total assets: 0.060*** (0.001) - Age: −0.023*** (0.001) - Observations: 3 291 433; R square: 0.452 (pooled baseline panel). - Policy recommendations to boost firm dynamism and productivity: - Reduce barriers to entry and competition, increase labor market flexibility, improve insolvency regimes, review credit guarantees that perpetuate low-productivity firms. - Address debt overhang through judicious debt resolution, corporate restructuring, targeted tax measures, and macroprudential tools. - Foster intangible investment via fiscal incentives, patent policy, competition in research grants, and openness to trade and FDI. ### Potential growth, demographic challenges, and policy actions - Potential growth declined after global crises in many Asian economies; decomposition attributes declines to capital, labor, and TFP contributions. - Working-age population expected to continue slowing in most Asian economies. - Policy actions to raise potential growth: - Support productive investment, boost productivity, increase female and older-worker participation, invest in physical and digital infrastructure, and enhance skills. ### Digitalization, ICT, fintech, e-commerce, and automation - Digital economy measurement: - Narrow OECD-based estimates: China 6 percent of GDP; Japan 8 percent; South Korea 10 percent. - Broader estimates (China Academy of Information and Communication): China 30 percent; Japan 46 percent; United States 59 percent. - ICT and robotics: - In 2016, about 1.8 million industrial robots worldwide; over 1 million in Asia. - Robot density (2016): Korea 63 robots per 1,000 manufacturing employees; Singapore 49; Japan 30; world average about 7.4. - Over 60 percent of industrial robots used in Asia; China has twice as many robots as Korea. - Fintech and e-commerce: - Cumulative Fintech equity funding about $28 billion in 2017, with two-thirds captured by China. - Firm-level evidence (WBES): e-commerce firms have higher TFP and export shares; in Asia, participation in e-commerce associated with more than a 30 percent increase in TFP and firms engaged in e-commerce export 50 percent more. - E-commerce sales share: China leads; Indonesia e-commerce sales increased fourfold between 2014 and 2017. - Digital public finance: - Closing half the gap to the digitalization frontier could raise import-VAT revenue by 0.6 percent of GDP in Asia and tariff revenue by 0.2 percent of GDP in Asia (estimates reported in source). - Automation impacts: - Cross-country sample (2010–14): robot penetration not significantly associated with net employment losses across 40 economies; within Asia, one more robot per 1,000 employees associated with a 0.26 percentage point decrease in manufacturing employment growth. - Japan prefecture analysis: a 1 percent increase in robot density corresponds to a 6 percentage increase in TFP and a 0.3 percentage point increase in employment growth in manufacturing (Table coefficients preserved). - Policy implications for digitalization and automation: - Invest in education and lifelong learning, training and re-training, SkillsFuture-type programs, and policies to increase female labor force participation. - Strengthen regulatory frameworks for Fintech (AML/CFT, consumer protection, cybersecurity), competition, data privacy, and taxation of digital platforms. - Use digital public services (digital ID, registries) to improve targeting and revenue collection while safeguarding privacy and inclusion. ### Modeling frameworks and data notes - Models used: - Ricardian Trade Model (RTM): multisector, multi-economy CGE model calibrated to 2014 WIOD with 35 sectors and 44 economies. - GIMF (Global Integrated Monetary and Fiscal Model): dynamic macro model with households, investment frictions, and bilateral trade. - Productivity channel calibration based on Ahn and others (2016) with coefficients: - α_out = -1.09872; β_out = -0.05205; α_in = -0.05205; β_in = -0.39589. - Firm-level dataset: - Orbis sample covering China, Japan, Korea, Malaysia, Philippines, Thailand; time span 2003–15. - Baseline panel regression reported with 3 291 433 observations (pooled results preserved). ### Key risks and medium-term challenges - Near-term and medium-term risks tilted to the downside. - Major risks: - Trade tensions (US-China tariffs announced June 15 and September 17, 2018, with further tariff possibilities). - Tighter global financial conditions and higher US dollar funding costs. - Corporate debt vulnerabilities and potential increase in corporate failures. - Home-grown risks: slower deleveraging in China, slower structural reforms in India, real-estate imbalances in Australia and Hong Kong SAR. - Climate change and natural disasters: 204 natural disasters recorded in 12 Pacific economies during 1980–2016, implying nearly a 50 percent chance of a country being hit in any given year and average damage of 14 percent of GDP. ### Consolidated policy recommendations (verbatim guidance and case-specific steps preserved) - General guidance: - Maintain expansion, contain risks, strengthen resilience to downside risks, and raise medium-term inclusive growth. - Tailor policy priorities across economies given cyclical positions and policy space. - Monetary and fiscal: - With low inflation and negative output gaps in most advanced economies in region, monetary policy should generally remain accommodative. - Fiscal policy should build buffers, support inclusive long-term growth, and reduce excessive external imbalances. - Specific fiscal guidance by case: - Support activity where domestic demand weak and fiscal space exists (Korea, Thailand, Singapore). - In China, gradual withdrawal of fiscal support to preserve policy space; composition to support rebalancing toward private consumption. - Where fiscal sustainability risks exist, embed near-term support in credible medium-term frameworks (Japan). - Continue fiscal consolidation where large fiscal imbalances entail macro risks (India). - Fiscal consolidation essential where current account deficits are large (Sri Lanka, Cambodia, Mongolia). - Build fiscal buffers where growth is strong (small states). - Exchange rate and capital-flow guidance: - In flexible exchange rate economies, allow depreciation while monitoring balance-sheet and inflation effects; support flexibility with appropriate policy tightening. - Use FX intervention only for disorderly markets; avoid fixing rates or conveying fixation. - Consider capital flow measures only in crisis or imminent crisis as contingency, not as substitutes for macro adjustment. - Financial-sector and macroprudential steps: - Reduce vulnerabilities to tightening financial conditions via micro- and macroprudential tools. - Guard against currency and maturity mismatches; ensure adequate FX hedging and hedging-counterparty quality. - Consider longer-term swap lines for banks where sufficient FX reserves exist. - Reign in excess credit growth (China) and support healthy bank balance sheets (India). - Supply-side and structural reforms: - Raise potential output via investment in physical and digital infrastructure, boost labor force participation (female and older workers), upgrade skills. - Alleviate infrastructure bottlenecks, strengthen business environment, and eliminate barriers to intraregional trade. *Italic: Source: IMF staff background paper (apd-reo-background-papers-1018).* ### 2.1 percent in 2019 as output gaps close and monetary ### apd-reo-background-papers-1018 - 2.1 percent in 2019 as output gaps close and monetary ### Global outlook and risks - World growth projected to decelerate to 2.1 percent in 2019 as output gaps close and monetary policy becomes less accommodative. - Emerging market and developing economies are projected to grow at 4.7 percent in 2018–19. - Risks around the baseline forecast are now seen as tilted more to the downside (near term) compared with being viewed as balanced in the April 2018 WEO. - Factors contributing to downside risk: - Weak growth outturns in the first half of 2018 in several large economies. - Moderation in high-frequency economic indicators. - Tighter financial conditions in vulnerable economies. - Risk of a further escalation of trade actions. - Geopolitical tensions and delays in addressing inequality and climate change. ### Regional developments (Asia) - GDP growth outturns in the first half of 2018 were softer than in 2017, especially in advanced economies; growth remained robust in emerging economies and broadly in line with expectations. - Headline inflation inched up in several economies due to higher oil prices and currency depreciation. - Export growth generally slowed in early 2018 as import growth stalled in advanced economies (especially Japan and the euro area). - Current account surpluses in the region narrowed, and many economies decreased their holdings of foreign reserves during 2018. ### Financial conditions and capital flows - Financial conditions tightened in some large Asian economies in response to US policy normalization, rising global trade tensions, and volatility in China and other large emerging economies. - Cumulative portfolio inflows in 2018 were far below 2016–17 levels due to large outflows during Q2 2018. - Nonresident sales of portfolio debt securities increased as the US dollar started to appreciate. - Asian equity indices and exchange rates were negatively affected by trade tensions; bond yields and spreads generally increased. - Some central banks raised policy rates; others intervened directly to support domestic currencies. - The recent volatility of Asian assets has been comparable in some respects to the taper tantrum but Asia has been affected much less than other regions and is more tied to developments in China. ### Selected country developments and key statistics - China - Investment growth slowed from 6 percent year-over-year in June to 4.1 percent in August. - Unemployment rose from 4.8 percent in June to 5 percent in August. - Headline CPI inflation remained contained at around 2 percent. - Credit growth slowed to 10.1 percent year-over-year in August, down from around 14 percent at end-2017. - Renminbi depreciated 7 percent against the US dollar since mid-June. - Equity prices declined substantially; money market rates have declined reflecting monetary easing. - Current account surplus continued declining. - Japan - Temporary contraction in Q1 2018; economy rebounded briskly in Q2 2018. - Headline inflation reached 1.3 percent year-over-year in August; core inflation reached 0.4 percent year-over-year in August. - India - GDP growth slowed to 6.7 percent in FY2017/18 but recovered to 7.7 percent year-over-year in Q1 2018. - Headline inflation averaged 3.6 percent in FY2017/18 (a 17-year low). - Inflation declined to 3.7 percent year-over-year in August from 4.2 percent in July. - Korea - Growth: 3.1 percent in 2017 and 2.9 percent in H1 2018. - Headline and core inflation: 1.4 and 0.9 percent year-over-year, respectively, in August. - Current account surplus: declined to 5.1 percent of GDP in 2017; narrowed further in H1 2018 by around 2.5 percentage points of GDP (year-over-year). - Australia and New Zealand - Australia: annual growth picked up to 3.3 percent in H1 2018; consumer prices rose 2.1 percent year-over-year in Q2 2018; core inflation 1.9 percent year-over-year. - New Zealand: GDP growth 2.8 percent year-over-year in Q2 2018 (up from 2.6 percent); headline inflation 1.5 percent year-over-year in Q2; core inflation 1 percent. - Hong Kong SAR - GDP growth remained strong in H1 2018; exports expanded 8 percent year-over-year in H1 2018 with a 5 percent increase in volumes. ### ASEAN and Southeast Asia highlights - Indonesia - Growth ticked down to 5.2 percent year-over-year in H1 2018 (from 5.1 percent in 2017). - Since January, rupiah depreciated by 10 percent against the US dollar; equity prices fell by 8 percent; 10-year government bond yields rose by 180 basis points. - Bank Indonesia raised its policy rate to 5.75 percent. - Thailand - GDP increased 4.8 percent year-over-year in H1 2018 (vs 4.2 percent prior semester). - Headline inflation 1.6 percent year-over-year in August; core inflation 0.8 percent. - Singapore - Real GDP grew 3.6 percent in 2017 and 4 percent year-over-year in H1 2018. - Headline inflation 0.7 percent year-over-year in August. - Malaysia - Real GDP grew 4.9 percent year-over-year in H1 2018 (vs 5.9 percent in 2017). - Headline inflation dropped from 3.8 percent in 2017 to 1.6 percent year-over-year in H1 2018. - Current account surplus: 2.7 percent of GDP in H1 2018 (from 3 percent of GDP in 2017). - Philippines - Growth: 6.7 percent in 2017 and 6.3 percent year-over-year in H1 2018. - Inflation rose to 6.4 percent year-over-year in August, exceeding the 2 to 4 percent target band. - Vietnam - Growth 7.1 percent year-over-year in H1 2018. - Inflation 4 percent year-over-year in August; core inflation 1.5 percent year-over-year. - Myanmar, Cambodia, Lao P.D.R., Brunei Darussalam - Myanmar: economy rebounded strongly in 2017/18 from 2016/17 slowdown. - Cambodia: growth around 7 percent; inflation 3 percent year-over-year in June. - Lao P.D.R.: indicators show slight moderation; inflation expected to remain at around 2 percent in 2018. - Brunei Darussalam: real GDP growth 1.3 percent in 2017; inflation -0.2 percent in 2017 and turned positive in early 2018. ### Other frontier and Pacific economies - Bangladesh - Real GDP growth in FY2018 projected to have accelerated to 7.5 percent. - Inflation eased to 5.5 percent by end of FY2018 after flood-related rises. - Nepal - Growth 6.3 percent in FY2017/18; inflation fell to 4.2 percent. - Sri Lanka - Gradual recovery under way after subdued growth in 2017. - Mongolia - Growth accelerated from 5.1 percent in 2017 to 6.3 percent year-over-year in Q1 2018. - Fiscal: 3.4 percent of GDP primary surplus in Q1 2018. - Pacific Island countries - Growth around 2.6 percent in 2017. - Fiji: growth jumped to 3 percent (recovery from Cyclone Winston). - Papua New Guinea and Timor-Leste: growth relatively subdued due to weak commodity prices. - Cyclone Gita caused widespread destruction in Tonga and Samoa. ### Regional outlook - Near-term outlook for Asia remains positive, supported by steady global momentum and broadly accommodative policies. - Downside risks are more pronounced than in the spring. - Asia continues to be the main growth engine of the world. *Italic: Source: IMF staff background paper (apd-reo-background-papers-1018).* ### 5.6 percent in 2018 (unchanged from what was ### apd-reo-background-papers-1018 - 5.6 percent in 2018 (unchanged from what was projected in April) ### Regional summary and headline projections - Asia: real GDP growth projected at 5.6 percent in 2018 (unchanged from April) and 5.4 percent in 2019, down by 0.2 percent from April; growth expected to moderate to lower rates in the medium term. - Inflation (Asia): projected to increase to 2.7 percent in 2018 and 2.9 percent in 2019, in line with higher commodity prices; expected to remain below inflation targets for several economies. - Current account (Asia): balances expected to narrow with higher oil prices. ### Key aggregate statistics (as reported) - Table 1 (Asia: Real GDP Growth): Asia 2018 = 5.6; 2019 = 5.4; Difference from April 2018 WEO = 0.0 (2018) and -0.2 (2019). - Table 2 (Asia: Consumer Prices): Asia 2018 = 2.7 percent; 2019 = 2.9 percent; Difference from April 2018 WEO = 0.0 (2018), -0.2 (2019). - Table 3 (Asia: Current Account Balance): Asia 2018 = 1.5 percent of GDP; 2019 = 1.4 percent of GDP; Difference from April 2018 WEO = 0.0 (2018), -0.3 (2019). ### China - Policy shift: authorities moved policy priority from deleveraging to stabilizing growth; State Council (mid-June) and Politburo (end-July) emphasized stabilizing domestic demand and managing deleveraging pace. - Growth: projections for 2018 remain unchanged; 2019 projection marked down from 6.4 to 6.2 percent reflecting impact of US tariffs announced on September 17 on a list of products imported from China worth $200 billion and offsetting policy action (mainly stimulating investment and credit). - Current account: surplus likely smaller than earlier forecast due to less favorable terms of trade. - Inflation (headline): projected to reach 2.2 percent in 2018 and 2.4 percent in 2019, slightly lower than before given weaker-than-expected recent outturns. ### Japan - Growth: projected to expand by 1.1 percent in 2018; expected to decelerate to 0.9 percent in 2019 as growth converges toward potential. - Inflation: revised up from 1.1 to 1.2 percent in 2018 mainly due to rising global energy prices; likely to remain below the 2 percent target given tepid wage growth. - Monetary policy: Bank of Japan strengthened commitments to accommodative policy in late July via forward guidance to maintain ultra-low policy rates for an extended period, considering economic uncertainties including the scheduled consumption tax hike in October 2019. - Current account: projected to moderate to 3.6 percent of GDP in 2018 because of softer-than-expected exports in the first half of 2018. ### India - Growth: projected at 7.3 percent in FY2018/19 and 7.4 percent in FY2019/20; revised down by 0.1 and 0.4 percentage points, respectively, due to higher oil prices and further monetary policy tightening. - Drivers: near-term growth underpinned by strengthening investment and robust private consumption. - Current account: deficit revised up to 3 percent of GDP in FY2018/19 due to rising oil prices and strong demand for imports. - Inflation (headline): projected at 4.7 percent in FY2018/19, above the mid-point of the Reserve Bank of India’s target band. - Medium-term: growth expected to rise to 7¾ percent reflecting ongoing structural reforms, including productivity-enhancing effects of the Goods and Services Tax. - Note: India's data are reported on a fiscal year basis (April 1 to March 31). ### Korea - Growth: GDP growth in 2018 projected to continue at around 2.8 percent. - Composition: private consumption likely to remain solid; government spending to expand more rapidly; investment contribution expected to fade as construction investment slows; exports expected to continue robust growth. - Inflation: projected to remain subdued amid persistent slack in the labor market. - External sector: current account balance projected to remain around 5 percent of GDP in 2018. ### Australia and New Zealand - Australia: approaching final leg of rebalancing after mining boom; recent strong employment growth and higher infrastructure spending supportive; output gap should be closing; inflation forecast to return to midpoint of target range within next three years; baseline assumes a soft landing in housing market. - New Zealand: growth expected to remain around 3 percent, driven by higher spending supporting infrastructure and structural reform; inflation should pick up gradually toward midpoint of 1 to 3 percent target range. ### Hong Kong SAR - Growth: 2018 revised up by 0.2 percentage points to 3.8 percent, mainly due to a stronger-than-expected fiscal impulse in FY2018/19 budget and favorable global trade environment. - Risks: high income inequality and an aging population could weigh on long-term prospects. ### ASEAN highlights - Indonesia: - Growth: projected at 5.1 percent in 2018-19. - Revision: downward for 2018–19 due to weaker-than-expected Q1 2018 outturn and tighter financial conditions. - Inflation: expected to rise slightly to near the center of the band (2.5 to 4.5 percent) reflecting rupiah depreciation and higher commodity prices. - Current account: expected to rise to 2.4 percent of GDP in 2018 from 1.7 percent in 2017. - Medium term: real GDP growth projected at 5½ percent. - Thailand: - Growth: momentum projected to continue in 2018 and 2019; 2018 real GDP growth revised up from 3.9 to 4.6 percent given stronger-than-expected H1 2018 outturn. - Inflation: projected to remain below the 2.5 percent midpoint of the target range (1 to 4 percent). - Current account: surplus projected to decline but remain very large. - Singapore: - Growth: on a strong cyclical upswing; growth expected at 2.9 percent in 2018, above potential, increasingly supported by domestic demand. - Inflation: subdued but expected to rise modestly. - Current account: surplus remains large. - Malaysia: - Growth: expected to moderate to 4.7 percent in 2018 and 4.6 percent in 2019, driven by lower public domestic demand contributions and some normalization in net exports and private consumption in 2019. - Inflation: headline inflation projected to drop to an average of 1 percent in 2018 because of temporary indirect tax changes and normalization of core inflation, then bounce back to 2.3 percent in 2019. - Philippines: - Growth: real GDP projected to grow at 6.5 percent in 2018, led by strong domestic demand. - Drivers: government infrastructure push and stable foreign direct investment expected to accelerate investment growth; private consumption projected to remain robust, underpinned by remittances and rising employment. - Monetary policy context: recent policy rate increases noted (text cuts off). *Italic: Source — IMF (apd-reo-background-papers-1018 PDF content provided).* ### 4.5 percent, inflation is projected to remain above the 4 percent target upper bound in 2018 and ### apd-reo-background-papers-1018 - 4.5 percent, inflation is projected to remain above the 4 percent target upper bound in 2018 and ### Outlook: ASEAN and selected economies - Inflation "4.5 percent, inflation is projected to remain above the 4 percent target upper bound in 2018 and stay in the upper half of the band (2 to 4 percent) during 2019–20." - Vietna m: - Growth projected at 6.6 percent in 2018. - Inflation forecast to rise to just under the 4 percent target, led by higher oil prices and gradual increases in administered prices. - Myanmar: - Growth expected to pick up toward the estimated potential of about 7 percent over the medium term, reflecting continued large foreign direct investment inflows and an improvement in public investment spending and efficiency. - Cambodia: - Medium-term growth projected to moderate to about 6 percent due to maturation in credit and real estate cycles and challenges in diversification and competitiveness. - Lao P.D.R.: - Growth expected to continue at about 7 percent in the near and medium term. - Brunei Darussalam: - Growth expected to accelerate to 2.3 percent in 2018, reflecting rejuvenation of oil fields and continuing support from infrastructure projects. ### Outlook: Other frontier and Pacific economies - Bangladesh: - Growth projected to remain robust at around 7.3 percent in FY2018/19, supported by strong investment. - Inflation rises to 6 percent, close to the Bangladesh Bank’s target, reflecting higher fuel prices. - Nepal: - Growth expected to moderate to 5 percent in FY2018/19, underpinned by robust private consumption from strong remittance inflows and increased government spending. - Inflation expected to rise to 5 percent as food prices normalize and activity approaches capacity constraints. - Sri Lanka: - Real GDP growth projected to normalize but remain below 4 percent in 2018. - Inflation projected to remain at around 5 percent, the midpoint of the Central Bank of Sri Lanka’s target band. - Mongolia: - Revised growth forecast from 5 to 6.2 percent in 2018, supported by stronger coal prices and expanding credit. - Inflation expected to remain below 8 percent. - Pacific Island countries: - Growth expected to decline to around 1.4 percent in 2018, mainly reflecting the earthquake in Papua New Guinea, then rebound to 4.1 percent in 2019. - Inflation expected to remain low in most economies. ### Risks (near term and medium term) - Overall: The balance of risks has shifted to the downside in the near term and remains tilted to the downside over the medium term. - Trade tensions between the United States and China: - U.S. announced on June 15 a list of products imported from China (worth $50 billion) that will be subject to a 25 percent tariff; China announced retaliation on a similar scale. - On September 17, the United States announced a further $200 billion in imports from China that would be subject to a tariff starting at 10 percent and rising to 25 percent by year-end; China announced tariffs on an additional $60 billion of U.S. imports. - The United States has suggested an additional $267 billion of goods may be hit with tariffs; separate proposals include tariffs on the automotive sector. - Potential impacts: undermine confidence, hurt financial markets, disrupt supply chains, discourage investment and trade, make tradable consumer goods less affordable, and boost inflation. - Tighter global financial conditions: - Higher-than-expected U.S. inflation could lead the Federal Reserve to tighten faster, reducing capital flows and raising funding costs for Asia. - Sudden deterioration of risk appetite, rising trade tensions, and political and policy uncertainty could also tighten conditions. - Two additional channels highlighted: - Higher US dollar funding costs: U.S. long-term policy actions pushing up US dollar funding costs; persistent deviations from covered interest parity in Asian economies affect hedging costs and drive US dollar funding rates. - Corporate debt vulnerabilities: Post-crisis corporate leverage in emerging Asia has risen, amplifying shocks as global liquidity tightens, interest rates rise, and growth slows; firms with foreign currency debt are especially vulnerable. - Home-grown risks: - China: focus on stabilizing growth may slow deleveraging and heighten medium-term risks. - India: slower progress on structural reforms could hold back medium-term growth. - Japan: market speculation of changes to accommodative policy could have global knock-on effects. - Country-specific domestic risks: high private sector leverage in Korea; inflated real-estate markets in Australia and Hong Kong SAR. - Climate change and natural disasters: - For the 12 economies in the Pacific during 1980–2016: 204 natural disasters recorded, implying a nearly 50 percent chance of a country being hit in any given year, and disasters caused damage averaging 14 percent of GDP. ### Medium and long-term challenges - Potential growth slowdown in several advanced economies in Asia due to slower working-age population growth, more moderate investment rates, and lower total factor productivity growth. - Gaps remain between emerging economies and advanced economies in income and a broader measure of welfare (income, consumption, health, inclusiveness, environment). - Four key challenges affecting medium- and long-term outlook: - Rebalancing: need to shift toward domestic demand as growth in advanced economies slows; trade tensions make rebalancing more important. - Population aging: slowing population growth and aging risk parts of Asia “growing old before they grow rich”; raising female labor force participation can help counter negative effects. - Productivity slowdown: decline in “corporate dynamism” with fewer young firms and more “zombie” firms since the global financial crisis; resolving “zombie congestion” and promoting startups could improve aggregate productivity. - Digitalization: digital innovation offers opportunities and challenges (e.g., robotics may boost productivity but risk structural unemployment); comprehensive policies needed to harness digital benefits. ### Policy recommendations - General: - Maintain current expansion, contain risks, strengthen resilience to downside risks, and raise medium-term inclusive growth. - Policy priorities differ across economies given diverse cyclical positions, structural constraints, and policy space. - Monetary and fiscal guidance: - With low inflation and negative output gaps in most advanced economies in the region, monetary policy should generally remain accommodative. - Fiscal policy should focus on building buffers, supporting inclusive long-term growth, and reducing excessive external imbalances. - Specific fiscal guidance: - Where domestic demand is weak, current account is large, and there is fiscal space, fiscal policy should support activity and rebalancing (Korea, Thailand, Singapore). - In China, gradual withdrawal of fiscal support warranted to preserve policy space and macro sustainability; composition should support rebalancing from investment to private consumption. - Where fiscal sustainability risks exist, embed near-term fiscal support in a credible medium-term fiscal framework (Japan). - Where large fiscal imbalances entail macro risks, continue fiscal consolidation with high-quality measures (India). - Where current account deficits are large, fiscal consolidation is essential (Sri Lanka, Cambodia, Mongolia). - Where growth is strong, build fiscal buffers for use in bad times (small states). - Resilience to financial tightening and capital flow reversals: - In flexible exchange rate economies, allow currencies to depreciate to new equilibrium levels while monitoring balance-sheet and inflation effects. - Support exchange rate flexibility with appropriate monetary or fiscal tightening consistent with price stability and fiscal sustainability. - Use foreign exchange intervention only for disorderly market conditions; avoid fixing exchange rates or giving the impression of fixing them. - Avoid changing subsidies and transfers so pass-through and relative price adjustments occur. - Consider capital flow measures only in crisis or imminent crisis situations, not as substitutes for macro adjustment. - Case-specific steps: - Economies with sluggish growth and inflation below target should avoid excessively tightening financial conditions and support growth by allowing exchange rate adjustment (Thailand). - Japan: sustained accommodative monetary stance desirable to reflate the economy. - Economies where inflation is above target: monetary tightening should continue to meet inflation targets (India, Philippines). - Economies with exchange rate pegs (Singapore) or dollarized economies (Cambodia) should use fiscal policy to support the peg. - Financial-sector measures: - Micro and macroprudential policies should reduce vulnerabilities to tightening financial conditions. - Supervisors should guard against currency and maturity mismatches, ensure appropriate foreign exchange hedging, and check hedging counterparties' quality. - As a contingency, economies with sufficient foreign exchange reserves could introduce longer-term swap lines with banks to ease liquidity concerns and facilitate orderly balance sheet unwinding. - Reining in excess credit growth (China) and supporting healthy bank balance sheets (India) to improve resiliency and boost medium-term growth prospects. - Supply-side and structural reforms: - Prioritize measures to raise potential output and productivity: invest in physical and digital infrastructure, boost labor force participation where aging drags labor supply, and enhance labor force skills. - Structural reforms should alleviate infrastructure bottlenecks, strengthen the business environment, upgrade human capital, and ensure access to opportunities for all. - Eliminate barriers to intraregional trade. ### Box 1 (Impact of tighter global financial conditions on Emerging Asia) — key points - Scenario constructed using the IMF’s Flexible System of Global Models: - U.S. long-term borrowing costs rise 50 basis points in 2018 and an additional 50 basis points in 2019 before gradually returning to baseline. - This reduces U.S. demand and raises sovereign and corporate risk premiums in other countries, depressing output and investment in Asia. - Monetary policy is eased in countries not constrained by the effective lower bound; fiscal policy works via automatic stabilizers. - Findings: - If Asian emerging markets can ease monetary policy, the negative impact is smaller (illustrated by red line in Figures 1.1 and 1.2 in source). - If Asian emerging markets are unable or unwilling to ease monetary policy amid capital outflows, investment falls by about 3 percent (Asian emerging markets, excluding China) relative to a benign baseline, and GDP is depressed by about 1 percent in the years following the tightening — more than double the impact when monetary policy can respond (illustrated by yellow line in Figure 1.2). - Asian advanced economies face negative spillovers from weaker output in Asian emerging markets via the trade channel; monetary policy leeway is assumed in advanced economies outside of Japan. *Italic: Content based on apd-reo-background-papers-1018 (PDF chapter/section).* ### 1. Real GDP, Asian Advanced Economies ### 1. Real GDP, Asian Advanced Economies ### Real activity and investment: Asian advanced and emerging economies - Charts presented: - Real GDP, Asian Advanced Economies (Percent difference, level) for 2017–2023 with axis labels from -0.8 to 0.0. - Real GDP, Asian Advanced Economies (excl. Japan) (Percent difference, level) for 2017–2023 with axis labels from -2.0 to 0.0. - Real Investment, Asian Advanced Economies (Percent difference, level) for 2017–2023 with axis labels from -2.0 to 0.0. - Real Investment, Asian Advanced Economies (excl. Japan) (Percent difference, level) for 2017–2023 with axis labels from -0.8 to 0.0. - Real GDP, Asian Emerging Economies (Percent difference, level) for 2017–2023 with axis labels from -1.2 to 0.0. - Real GDP, Asian Emerging Economies (excl. China) (Percent difference, level) for 2017–2023 with axis labels from -2.5 to 0.0. - Real Investment, Asian Emerging Economies (Percent difference, level) for 2017–2023 with axis labels from -4.0 to 0.0. - Real Investment, Asian Emerging Economies (excl. China) (Percent difference, level) for 2017–2023 with axis labels from -0.8 to 0.0. - Common scenario annotations across charts: - +Sovereign and Corporate Risk Spreads - + Interest-rate Constraint in Asian EMs - Note: EMs refers to Emerging Markets. Source: IMF Staff Calculations. ### Key statistic presentation conventions used in figures - Time coverage: 2017, 2018, 2019, 2020, 2021, 2022, 2023. - Percent-difference levels explicitly shown on each axis (no aggregation or averaging applied). --- ### Cross-cutting theme: Sovereign and corporate risk spreads; interest-rate constraints in Asian EMs - Figures explicitly overlay the impacts of: - Sovereign and Corporate Risk Spreads. - Interest-rate Constraint in Asian EMs. --- ### Box 2. What Do Deviations from Covered Interest Parity and Higher US Dollar Funding Costs Mean for Asia? ### Findings on cross-currency basis and US dollar funding costs - The ongoing US monetary tightening has put upward pressure on US dollar funding costs. - The cross-currency basis is an important component of US dollar funding cost; under covered interest parity, the basis is nil. - A negative (positive) basis implies a premium (a discount) the borrower of the US dollar pays to the lender. - Some Asian economies exhibit persistent deviations from covered interest parity that have not closed since the global financial crisis. - Economies with consistently positive deviations: Australia, New Zealand, and India. - Economies with consistently negative basis: Japan, Korea, and Hong Kong SAR. - Economies with mostly small deviations: Singapore, Thailand, and the Philippines. - Empirical charts: - Figure 2.1: One-Year Cross-Currency Basis (Basis points, quarterly average) (time series spanning 2000:Q1–2018:Q3) with axis showing points from -200 to 200. - Figure 2.2: One-Year Cross-Currency Basis (Basis points, monthly average) (time series spanning 2008:M1–2018:M1) with axis showing points from -100 to 40. - Figure 2.3: Net International Investment Position and One-Year Cross-Currency Basis (Average, 2014-17) plotting One-Year Cross-Currency Basis (basis points) against Net International Investment Position (in billions of US Dollars) with axis ranges shown (basis points from -60 to 20; NIIP from -1000 to 3000). ### Drivers of covered interest parity deviations (enumerated) - Banks’ counterparty credit risks and constraints on the US wholesale funding market. - Negative cross-currency bases observed during the global financial crisis in 2007–08 and the euro area sovereign debt crisis in 2011–12. - Since the beginning of 2014, the non-zero cross-currency basis has persisted despite stable credit risk indicators and no significant US dollar funding shortages among banks. - Structural factors among different sources of hedging demand and supply for US dollars. - International imbalances between investment demand and funding supply across currencies open covered interest parity deviations. - A negative relationship is found between net foreign assets and the cross-currency basis (proxying hedging demand by net international investment positions). - Example: in Korea, life insurance companies’ hedging demand is estimated to exceed banks’ hedging demand by five-fold. - Balance sheet constraints for financial intermediaries. - Regulatory reforms since the global financial crisis, such as tighter capital requirements, have increased balance sheet costs imposed on global banks that often act as arbitrageurs in closing covered interest parity deviations. ### Policy implications and recommendations - Policies to ensure adequate availability of US dollars could help mitigate the impact of a tightening in global financial conditions on financial stability and growth: - Central banks’ swap lines, already in place for some economies to provide foreign exchange liquidity in times of stress, could be tapped to facilitate orderly unwinding of foreign exchange positions. - Domestic capital market development for alternative funding instruments could protect domestic borrowers from being overly reliant on US dollar funding. ### Sources and references cited within box - Sources: Bloomberg; IMF Staff Calculations. - Related literature cited in text (no additional details provided beyond names and years). --- ### Box 3. The Turning Tide: How Vulnerable Are Asian Corporates? ### Key findings on corporate debt and vulnerabilities - Global corporate debt has more than doubled over the past decade. - In emerging Asia, corporate debt increased to about 114 percent of GDP in 2017:Q4 from 71 percent in 2007:Q4. - Corporate debt levels are generally higher in Asia than other regions, particularly in China, Hong Kong SAR, and Singapore. - Micro data indicate: - The share of foreign currency (mainly US dollar) debt has declined over the past decade. - The average maturity of foreign currency debt has decreased, while that of local currency debt has increased. - Altman Z-score evidence: - Asia’s nonfinancial corporate sector is healthier than in the runup to the Asian financial crisis, but corporate vulnerabilities are higher in some Asian economies, including Hong Kong SAR, Indonesia, the Philippines, and Singapore. - Figure 3.1: Nonfinancial Corporations Credit (Percent of GDP) plotted for multiple economies and dates including 1996:Q4, 2007:Q4, 2017:Q4. - Figure 3.2: Fraction of Vulnerable Nonfinancial Corporates (Fraction of firms with Z-scores outside the safe zone) with country-level time series including China, Hong Kong SAR, India, Indonesia, Korea, Malaysia, Philippines, Singapore, Sri Lanka, Taiwan POC, Thailand. ### Quantified vulnerability to shocks - Regression analysis result: - The probability of financial distress (inverse of the Z-score) increases with local currency depreciation vis-à-vis the US dollar and with higher global interest rates. - Example quantification: a currency depreciation of 15 percent would reduce the implicit bond rating capturing the probability of default by one notch (e.g., A to A-). - Intuition: - Currency depreciation increases firms’ debt burden, especially for firms borrowing heavily in foreign currencies. - Tightening of US monetary policy significantly increases corporate sector vulnerability in Asia. ### Policy recommendations - Monitor vulnerable firms, especially systemically important ones, and closely monitor banks and other sectors linked to them. - Fill data gaps on corporate sector finances, including foreign currency exposures. - Use macroprudential and microprudential policies to limit buildup of foreign exchange balance sheet exposures and contain excessive increases in leverage. - Asian emerging markets should prepare for an increase in corporate failures and, where needed, reform corporate insolvency regimes. ### Table: Altman's Z-score and Bond Rating (verbatim ranges) - Z-score > 8.5 = AAA - 7.60 - 8.15 = AA+ - 7.30 - 7.60 = AA - 7.00 - 7.30 = AA- - 6.85 - 7.00 = A+ - 6.65 - 6.85 = A - 6.40 - 6.65 = A- - 6.25 - 6.40 = BBB+ - 5.85 - 6.25 = BBB - 5.65 - 5.85 = BBB- - 5.25 - 5.65 = BB+ - 4.95 - 5.25 = BB - 4.75 - 4.95 = BB- - 4.50 - 4.75 = B+ - 4.15 - 4.50 = B - 3.75 - 4.15 = B- - 3.20 - 3.75 = CCC+ - 2.50 - 3.20 = CCC - 1.75 - 2.50 = CCC- - Z-score < 1.75 = D - Source: Altman (2005). --- ### Box 4. Potential Growth in Asia: Recent Dynamics and Drivers ### Findings on potential growth dynamics - Potential growth declined following the global financial and euro area crises in a set of large advanced and emerging economies in the Asia and Pacific region. - The analysis decomposes potential growth into contributions from capital (K), labor (L), and total factor productivity (TFP or Y), using IMF country team estimates aligned with IMF (2018b) and (2018c). - Data sources: - Share of labor and capital stock from the 2014 Penn World Tables, extended for 2015–17 using gross fixed capital formation. - Potential employment constructed using working-age population and trend labor force participation. - Working-age population data from United Nations World Population Prospects; trend participation from ILO. - For China and India, participation assumed constant due to data constraints. ### Advanced economies subgroup (six economies) - For a group of six advanced economies in Asia, the decline in potential growth witnessed after the global financial crisis has been partly reversed in recent years. - Each factor of production (capital, labor, TFP) has seen a modest improvement. - Japan notably contributed to the improvement due to: - Less-negative working-age population growth. - An increase in trend labor force participation. - An uptick in investment and capital-stock growth. ### Excluding Japan - When Japan is excluded, potential growth shows a different picture: - Waning contribution from labor associated with slowing working-age population growth. - Moderating investment rates and lower TFP growth have weighed on potential growth. ### Emerging markets subgroup - For a group of six emerging economies in Asia, there has been a marked slowdown in potential growth since around 2010. - The slowdown is explained by lower contributions from capital, labor, and total factor productivity. ### Projections and policy implications - The working-age population is expected to continue slowing in most Asian economies (Figure 4.4 shows working-age population growth rates for Asian AE, Asian AE excl. Japan, Asian EM, Asian EM excl. China). - Policy actions needed to sustain or increase potential growth include: - Measures supporting productive investment. - Policies encouraging increases in productivity. - Policies better incorporating female and older workers into the labor force. --- ### Box 5. Sustainable Development and Economic Welfare in Asia ### Key findings on income convergence and welfare - Asia’s GDP per capita increased from 8 percent of US GDP per capita in 1950 to 21 percent in 2016 (Figure 5.1). - Catch-up has been more pronounced since the Asian financial crisis in 1998. - Based on current projections, GDP per capita in Asian economies could reach 36 percent of US GDP per capita by 2030. - Broader measure of economic welfare (a function of selected Sustainable Development Goal indicators) shows a larger gap with the United States: - Welfare gap with respect to the United States is 84 percent. - Income gap is 78 percent. - Under current projections, convergence in economic welfare is likely to lag income per capita improvements: - Welfare level in the median Asian economy projected to improve from 16 percent of the US level to about 28 percent by 2030. - Projected welfare improvements driven primarily by higher income and life expectancy. ### Components and policy guidance - The welfare measure used includes income, consumption, life expectancy, income inequality, and greenhouse gas emissions. - Policy recommendations to accelerate welfare convergence: - Achieve more balanced, inclusive, and environmentally sustainable development outcomes. - Incorporate Sustainable Development Goal priorities into national development plans. - Account for costs associated with policies to improve sustainable development by improving spending efficiency and mobilizing higher domestic revenues and other resources, including from the private sector. - For lower-income Asian countries with large sustainable development needs, additional concessional financing from development partners would likely be required. ### Numerical and reference notes appearing with figures - The Greenhouse gas cost per unit of consumption of advanced countries is 1.4 percent (2017 and 2030). - The Gini coefficient of advanced countries is 41.6 percent (2017) and 42.4 percent (2030). - Life expectancy in advanced countries is 79 years (2017) and 81.5 years (2030). - Figure 5.2 shows components of welfare with the black line the median for Asia and dotted lines for medians of high-, middle-, and low-income Asia; ratio of consumption to income is expressed in terms of advanced economy comparator. *Source: IMF Staff Calculations as presented in apd-reo-background-papers-1018 - 1. Real GDP, Asian Advanced Economies.* ### Box 6. Increasing Female Labor Force Participation in Asia ### Box 6. Increasing Female Labor Force Participation in Asia ### Overview - Female labor force participation in Asia increased from 48 percent in 1990 to 54 percent in 2016. - Male labor force participation has dipped slightly over the same period. - The rise in female participation occurred despite downward pressures related to population aging. - Drivers include strong growth, expansion of the service sector, higher education levels, lower labor market rigidities, and family-friendly policies (for example, childcare provision and maternity job protection). ### Cross-country patterns and magnitude of change - Several economies with low female labor force participation before the Asian financial crisis recorded significant (above 10 percentage points) improvements during 1996−2016: Australia, Singapore, Malaysia, Hong Kong SAR, and New Zealand. - Korea marginally improved on historically stagnant or declining participation rates. - Laggards include India, Indonesia, the Philippines, and Sri Lanka; the gap between better and poor performers in Asia has been increasing over time. - Median female labor force participation figures referenced for groups: female median (16 Asian countries) and female median (21 advanced economies, excluding Asia). ### Age composition and decomposition - On average, prime-age female workers are participating at higher rates than before. - Decomposing changes shows that aging has reduced the share of prime-age workers in the overall population, thereby reducing the size of the overall improvements in women’s participation. - Changes in female labor force participation during 1997−2016 are decomposed into: average aging effect, average growth effect, and an unexplained component. The unexplained component remains large even after accounting for aging and growth. ### Key drivers (age 25–54) - Structural factors most closely associated with higher female labor force participation: - Relative service sector employment (versus industrial sector) - Years of schooling (higher education levels) - Other factors with positive but smaller average effects given Asia’s diversity: - Openness (lag of trade as % of GDP) - Pace of urbanization (not shown) - Part-time employment (% of total) - Index of maternity protections - Higher cost of automation (relative price of investment) is positively linked with female participation because it reduces the incentive to automate. - Automation (routinization) is generally detrimental to female labor force participation, though exposure varies across economies (for example, higher agriculture shares reduce exposure). - Maximum severance pay and advance notice periods, and initial routinization measures are among labor-market institutional variables considered. ### Policy implications and recommendations - Asian policymakers should aim to adopt family-friendly policies and continue to pursue reforms that support growth. - Family-friendly policies and structural characteristics proxied by an index of maternity protections and part-time employment that are tailored to address specific constraints faced by women play a positive role. - Data gaps, the large informal sector, and limited institutional capacity to enforce legally mandated provisions in some Asian economies dampen the overall effects of policies. - Labor market rigidities that make it difficult for businesses to let go of unproductive workers can have a chilling effect on employers’ decisions to hire female workers. - Bringing more women into the workforce could help counter shrinking labor forces from population aging, raise growth via expansion of the total labor force, reduce income inequality, diversify the economy, and boost firm profitability. *Source: apd-reo-background-papers-1018 - Box 6. Increasing Female Labor Force Participation in Asia* ### 1. Global Import Volume Growth ### 1. Global Import Volume Growth ### Asia’s trade integration and global value chains (GVCs) - The rise of GVCs has been particularly pronounced in China, Korea, Japan, and the Association of Southeast Asian Nations (ASEAN). - All Asian subregions (East Asia, Southeast Asia, South Asia, Pacific-Oceania, and the rest of Asia) trade more intensively with other subregions than within themselves, except for East Asia. - Asia’s intraregional goods trade: - Accounts for roughly 20 percent of global trade. - Nearly 60 percent of the region’s total trade is intraregional, i.e., within Asia. - Asia’s intraregional goods trade is mostly in manufactured intermediate goods, reflecting the region’s participation in GVCs. - Asia is second only to Europe in the size of its intraregional trade and is on an upward trajectory. - Asia’s trade openness and participation in GVCs has plateaued and in some cases (such as China) declined since the global financial crisis. - Possible factors: smaller declines in trade costs, reshoring, higher obstacles to cross-border investment, maturation of GVCs. - In China, lower foreign value added embodied in exports (backward linkages) is attributed in part to onshoring and increased domestic production of intermediate inputs. - Low-income Asian economies have, with a few exceptions, not participated significantly in GVCs. - Intermediate exports from Asia are largely medium/high-tech manufacturing from advanced and emerging economies. - Vietnam is noted as a relatively new emerging market gaining market share in electronics. - Other low-income Asian economies are concentrated in low-tech light manufacturing (for example, Bangladesh, Cambodia and Myanmar). - India, Hong Kong SAR, the Philippines, and Singapore stand out as GVC-related service exporters. ### Measures of GVC participation and trade composition - GVC participation is defined as the sum of foreign value added in domestic exports (backward participation) and domestically produced intermediates used in third economies (forward participation) expressed as the ratio to an economy’s gross exports. - Selected GVC and value-added observations (as presented): - Advanced and emerging Asia drive intermediate goods trade with limited participation by Asian low-income countries. - Emerging Asia (particularly China) is also the dominant provider of final goods abroad. - Asia intermediate and final exports composition, 2016: - Intermediate exports and final exports broken down by categories such as Food and beverages; Industrial supplies; Fuels and lubricants; Parts and accessories of capital goods; Transport equipment; Durable consumption goods; Non-durable consumption goods (figures provided in source). ### Trade openness and services trade - Services constitute some two-thirds of global GDP and employment, and a quarter of global trade (nearly half of global trade measured on a value-added basis). - Asia’s shares of global service imports and exports have both increased; Asia’s service trade is currently second only to the euro area. - Asia has experienced growth across all service sectors, particularly in travel and telecommunications/business services, though service export levels remain relatively low compared to goods exports due to higher trade barriers and trade costs. - Gravity-model results: - Asia overtrades in goods (exports more than predicted by the model) by a wide margin relative to most other regions, though the degree of overtrading has decreased since the global financial crisis. - Asia trades more within the region than predicted by the model, reflecting GVC impacts. - For services, Asia undertrades by a wide margin relative to the euro area and is in a similar position to Latin America and developing Europe; intraregional trade in services in Asia has fallen behind. ### Exposure to external demand and destinations of value added - The importance of GVCs and intermediate goods trade linked to final demand in other regions leaves Asia exposed to external demand conditions. - The majority of export exposure in value-added terms is still to non-Asian advanced and emerging economies (figure-based evidence provided in source). ### Impact of trade tensions: channels, scenarios, and modeled outcomes - Three channels of impact modeled: 1. Trade channel: higher tariffs depress relative demand for imports, compounded by investment and consumption responses to permanent income losses. 2. Confidence effects: short-term depressed business confidence lowers private business investment in 2018 and 2019 before dissipating; magnitudes linked to economies’ trade openness relative to that of the United States. 3. Financial effects: higher corporate spreads peak in 2019 and further drag on private business investment in 2019 and 2020; magnitudes linked to historic episodes and current credit ratings. - Tariff scenarios (three building layers): 1. Baseline scenario (measures already implemented and those announced on September 17, 2018): - 25 percent tariff on steel and a 10 percent tariff on aluminum imposed by the United States against a wide range of economies (mainly China in Asia), with retaliation against a diverse array of US goods. - 25 percent tariff between the United States and China on $50 billion of goods. - On September 17, the United States announced a further $200 billion in imports from China subject to a tariff starting at 10 percent and rising to 25 percent by year’s end. - China is assumed to retaliate through tariffs varying from 5 to 15 percent in the fourth quarter of 2018 and rising to a maximum of 25 percent in 2019 on $60 billion of imports from the United States. 2. Escalation scenario: - Assumes the United States imposes a 25 percent tariff on all goods imports from China (worth over $500 billion). - Assumes China retaliates against all goods imported from the United States with a uniform tariff of 25 percent. 3. Automobile sector scenario: - Adds a 25 percent US tariff on imports of cars, trucks, and automobile parts, with retaliation by economies (especially Japan and Korea for Asia) beyond just the automobile sector to match the value of US tariffs. - Example calculation: if automobiles represent around 15 percent of US imports of final goods, the impact of a 25 percent tariff on the automobile sector would be calculated as the impact of a 3.75 percent tariff on all imported final goods. - The auto sector scenario references an estimated value of about $350 billion for affected imports. - Timing note: - Effects of tariffs in 2018 only cover six months for the baseline tariffs (or three months in the case of China and US tariffs related to the $200 billion of China’s goods). - Escalation and automobile sector tariff scenarios are assumed to be levied starting in 2019. - Modeling assumptions: - Governments and central banks follow standard reaction functions and do not make extraordinary accommodative policy changes. - Japan is assumed unable to ease conventional monetary policy because of the zero lower bound; additional unconventional measures by the Bank of Japan would roughly halve the projected Japanese real GDP decline. - Short-term limited ability for firms to substitute between intermediate inputs from different economies; higher substitutability over the long term. - Confidence and financial market effects are uncertain and provided as separate layers. ### Modeled macroeconomic impacts and illustrative results - High-frequency evidence on trade policy uncertainty: - A daily news-based trade policy uncertainty index was constructed for April to July 2018. - Abnormal equity market returns are negatively impacted by shocks to trade policy uncertainty, with heterogeneity across economies. - Emerging Asia, excluding China, experienced four times the negative impact on abnormal returns as the United States and two times that of Japan. - Economies that trade more (total exports plus imports as a percent of GDP) with China and the United States have been worst off in terms of equity market reaction. - GIMF + RTM model findings summary (illustrative): - The effects of recently enacted tariffs and retaliation are small but material, especially for China. - GDP losses would rise substantially if additional tariffs are implemented, particularly if business confidence and financial markets are affected. - For most countries, the output effects of tariffs fade after a few years, but lasting effects could be substantial in China, Korea, and the United States. - If all channels were in play (tariffs enacted/proposed plus confidence and financial effects): - Peak GDP losses of 1.6 percent in China. - Peak GDP losses close to 1 percent in the United States. - Peak GDP loss for Asia as a whole would be 0.9 percent. *Source: apd-reo-background-papers-1018 - 1. Global Import Volume Growth (IMF staff calculations and text as provided in the source PDF).* ### 1.6 percent of GDP are also broad-based in ### 1.6 percent of GDP are also broad-based in ### Trade-tension impacts on China and Asia - Under the full scenario (baseline, escalation, and auto sector tariffs with confidence and financial market effects): - China: real GDP falls by 1.6 percent (green lines); if automobile sector tariffs do not materialize, China would be worse off under the escalation scenario alone, with real GDP being 1.8 percent lower (red lines). - Investment in China is hit by 4.6 percent. - Labor income in China falls, reducing employment and the real wage for still-employed workers. - Consumption in China falls by 1.9 percent. - Peak responses are generally from 2020 (note: for China and Australia peaks are from 2019). ### Peak impacts on real GDP across Asia (Figure 8) - China: largest losses through the trade channel. - Korea: real GDP growth could fall by around 0.9 percent. - India, Indonesia, Japan: real GDP growth could fall by around 0.7 percent. - Philippines: real GDP growth could fall by around 0.6 percent. - Australia, Malaysia, Thailand: real GDP growth could fall by around 0.5 percent. - New Zealand: real GDP growth could fall by 0.3 percent. - Some ASEAN economies (Indonesia, Malaysia, the Philippines, and Thailand) suffer a financial effect on the order of 0.4 to 0.6 percent of GDP. - India has a strong negative financial effect stemming from its own financial vulnerabilities and weaker corporate credit rating. ### Trade volumes and channels (Figure 10 and related text) - Trade volumes weaker in Asia: - Japan: led by Japan at 2.2 percent (trade volumes weaker by 2.2 percent). - China: more than 1.5 percent weaker. - Korea: almost 1.2 percent weaker. - Rest of Asia: around 0.2 to 0.6 percent lower. - Drivers include spillovers from confidence and financial market impacts, trade diversion, and real exchange rate movements. - Broad-based US dollar appreciation and broad-based Chinese renminbi depreciation lead to cheaper imports for economies outside the trade war, driving up import volumes and partly offsetting negative export spillovers. ### Sectoral and labor reallocation effects - Total labor reallocation remains limited: - Ranges up to 0.3 to 0.4 percent of employment in China, Korea, Japan, and the United States (Figure 11a); even smaller for other Asian economies. - Larger sectoral displacements in escalation and automobile tariff scenarios: - Escalation scenario: China’s electrical equipment and the US agricultural sectors shed more than 2 percent of their initial employment. - Metals sector often most affected due to steel and aluminum tariffs. - Automobile tariff scenario: transport sector labor shedding reaches around 12 percent of sectoral employment in Japan, 7 percent in Korea, and 8 percent in Taiwan Province of China (Figure 11b). - In the escalation scenario many US sectors raise their prices after being sheltered by foreign competition and thereby lose competitiveness in international markets, particularly metals (Figure 12). ### Trade diversion and supply-chain channels (Figure 13) - Two channels in the Ricardian Trade Model (RTM): - Trade diversion: US imports are substituted away from China toward other suppliers. - Supply-chain channel: all economies export less to China as China’s demand falls. - Sector-level examples: - Lower Chinese exports of electrical equipment to the United States are substituted by other suppliers. - Asia value chains for electrical equipment supply less to China, reducing exports across the chain. ### Role of trade liberalization and policy implications - Policies aimed at reducing trade costs and reinvigorating trade remain important given trade tensions and unfavorable productivity trends. - Investment in tradable sectors contributes to faster productivity growth; external factors account for almost half of GDP per capita growth (Figure 14). - Empirical evidence: lowering average tariffs significantly increased the region’s growth from 1990 to 2007, but its statistical significance has declined over 2005–17 (Figure 15). - Reducing trade costs—including tariffs and a broad range of other factors that drive a wedge between exporter producer prices and importer consumer prices—should be a priority. ### Trade costs and non-tariff barriers (NTBs) - Overall trade costs have fallen but remain high in Asia, especially in services (Figure 16). - Policy-related NTBs account for the vast majority (60 to 90 percent) of trade costs; tariffs and natural trade barriers account for the remaining 10 to 30 percent (Duval and Utoktham 2011). - Overall trade costs for manufactured goods have fallen less than effective tariff rates, suggesting substantial remaining policy-related NTBs. - Asia shows high service-sector barriers to foreign entry and competition and foreign investment restrictions (Cerdeiro and Nam 2018). - For both emerging and advanced economies, Asian economies are more restrictive than their peer groups across several NTB trade indicators; emerging market Asia’s service trade restrictiveness is higher than the EU-20 and Latin America in all dimensions. *Source: IMF staff calculations and IMF staff estimates (apd-reo-background-papers-1018).* ### 5. Tariffs in Asia Pacific ### 5. Tariffs in Asia Pacific ### Decomposition of trade costs and remaining barriers - Trade costs include multiple components beyond tariffs: direct tariff costs (0-10%), natural trade barriers (geography & cultural distance) (10-30%), availability/use of ICT services (10), business environment (10), liner shipping connectivity (25), direct cost of trade procedures (1), other factors (e.g., indirect cost of trade procedures; currency rates; non-tariff measures (SPS, TBT)) (50), and policy-related non-tariff costs (60-90%). - Source for decomposition: Duval & Uthoktham (2011). - The rapid rise of GVCs in Asia increases complementarities between trade and investment; barriers to foreign investment in several sectors remain high in Asia. - Ease of starting a foreign business in Asia is "much more difficult than in the G7 and slightly more difficult than the G20 average" (World Bank’s Ease of Doing Business index; OECD’s Regulatory Restrictiveness Index). - In terms of FDI regulatory restrictiveness, restrictions on investment in agriculture, manufacturing, and services are higher in the region compared to Europe, North America, and Latin America. ### Barriers to services trade and FDI - Services trade and services FDI restrictions are particularly important given the already strong liberalization of goods. - The paper highlights high restrictions in services where liberalization could encourage services trade—especially professional services, wholesale, and retail. - The OECD Service FDI Regulatory Restrictiveness Index shows higher restrictiveness in some Asia economies for Wholesale & retail, Media, Telecommunications, Professional services (index scale: higher = more restrictive). ### Liberalization scenarios analyzed - Three illustrative trade liberalization and integration scenarios (progressively more open): - (1) "China Opens": China eliminates goods tariffs and reduces services NTBs for all trading partners. - (2) "Within Asia": All of Asia eliminates goods tariffs and reduces services NTBs for its trading partners in Asia. - (3) "Asia Opens": All of Asia eliminates goods tariffs and reduces services NTBs for all its trading partners. - Variant considered for scenario (3): "Asia Opens + FDI" — FDI liberalization defined as reducing an economy’s restrictive practices to at least the current average restrictiveness in each sector for a broad cross-section of economies. - Calibration specifics: - Goods tariffs are reduced to the lowest value among Asian economies for each sector; for simplicity most sectors reduced to zero. - Services NTBs are reduced by either 10, 20, or 30 percent, based on whether economies are considered to have lower, average, or higher NTBs. - Countries classified as low NTBs: Australia, Cambodia, Hong Kong SAR, Japan, Korea, Lao PDR, Mongolia, New Zealand, Singapore. - Countries classified as high NTBs: India, Indonesia, Philippines. - Remainder of Asia considered average NTBs. ### Channels and modeling approach - Two primary channels analyzed: - Trade channel: effects from cutting tariffs or NTBs, domestic and spillover effects via trade linkages. - Productivity channel: domestic productivity increases per Ahn and others (2016) and spills over via demand and real exchange rate movements. - Models used: - RTM (Regional Trade Model): captures long-term trade channel, GVCs, sectoral details, disaggregated sectors, and compounding productivity effects across sectors and borders. - GIMF (Global Integrated Monetary and Fiscal model): illustrates macroeconomic dynamics and transmission channels; less able to capture full GVC and cross-sectoral productivity amplification compared to RTM. - A potential third drag on productivity gains is restrictive FDI regimes; liberalization of FDI can amplify productivity and export gains in certain sectors. ### Quantified impacts on trade, structure, and income - Under scenario 3 ("Asia Opens") with FDI liberalization: - Exports of services rise by 17 percent (world trade; change from baseline). - Exports of goods rise by 9 percent (world trade; change from baseline). - Asia’s long-term GDP per capita: - Rises by 7 percent under Asia opening scenarios without FDI liberalization. - Rises by 15 percent if FDI is also liberalized. - World trade effects: - Asia creates new demand and substitutes some trade formerly taking place outside the region, but overall global trade increases. - Gains are larger when Asia opens to the world (scenario 3 and the FDI variant) compared to more limited regional opening. - Sectoral shifts: - Services experience a marked reduction in trade costs as NTBs are relaxed; some goods sectors (e.g., agriculture and food processing) also see large reductions in effective tariffs. - Labor reallocates into services away from agriculture at the regional level. - Several economies and sectors develop new comparative advantages when trade and FDI restrictions are relaxed (examples given in the analysis): - Indonesia: comparative advantage in ICT and professional services with FDI liberalization. - China: potential export driver in transport with FDI liberalization. - Korea: potential export growth in agriculture with FDI liberalization. - Australia: could export more services and become less dependent on mining exports. - India: could develop comparative advantage in professional services by loosening FDI restrictions. - Trade creation tends to outweigh trade diversion, especially for Asian economies; imports surge in sectors where trade costs decline most (e.g., agriculture for India and Korea; ICT, financial, and professional services for all Asian economies). - Productivity channel effects dampen some import increases and boost exports in sectors that open, particularly where FDI regimes are relatively open. ### Policy implications and adjustment needs - Further trade liberalization and regional integration—especially liberalizing trade in services and FDI—should help offset rising protectionism and generate growth and jobs. - Domestic policies are needed to smooth transitions and address trade-related labor dislocations, particularly in protected and sensitive sectors such as agriculture. - Negotiating reforms will likely take many years; compensatory and supportive domestic policies can help ensure broad-based sharing of gains and overcome resistance from vested interests. *Source: IMF staff estimates and analysis in "5. Tariffs in Asia Pacific" (apd-reo-background-papers-1018).* ### 1. China's Export Share by Sector ### 1. China's Export Share by Sector ### Key quantitative rebalancing and productivity findings - Rebalancing in Asia: Value Added by Sector (Figure 27) — percent of total: - Primary: 16.2, 14.4, 13.3 - Secondary: 31.6, 33.8, 33.3 - Tertiary: 52.2, 51.8, 53.4 - Note: values correspond to Baseline, Asia opens, Asia opens + foreign direct investment (in that order). - Global Changes in Real Wages (Figure 29) — percent deviation relative to baseline: - Reported values: 11.4, 10.4, 10.0, 33.0, 33.3, 33.6, 55.7, 56.2, 56.4 - Liberalization reduces prices of goods and services outside of Asia by 5 to 6 percent on average, leading to higher real wages globally. - Labor reallocation impacts (Figure 30): - The secondary sector in the United States and Europe sheds around 1 million workers, or more than 1 percent of initial employment. - Charted percent changes for primary/secondary/tertiary sectors are shown relative to initial employment for European Union, United States, and two scenarios (Asia opens; Asia opens + foreign direct investment). ### Channels of long-term GDP gains and transmission - Long-term GDP gains are mainly attributable to the productivity channel, with changing trading patterns having greater impact during transition phases. - Goods trade generates a larger share of aggregate productivity gains than services trade because services account for 20 to 30 percent of total trade, despite similar percent productivity gains between goods and services over time. - Within goods: - Productivity gains accrue more to final goods than intermediate goods because tariff cuts in final goods are generally much larger (example: China). - However, the tariff elasticity of output favors intermediate goods over final goods; cuts in intermediate goods tariffs increase GVC participation for much of emerging Asia. - Services liberalization considerations: - Most economies have much greater scope for reducing NTBs than tariffs. - Reducing NTBs is less costly for government finances than lowering tariffs because NTB reduction forgoes a cost rather than revenue. - Both goods and services liberalization impose transition costs through removal of protection and associated firm and employment adjustment. ### Effects of Foreign Direct Investment (FDI) liberalization - Scenario 3 reduces FDI restrictions to the global average; FDI can increase gains in real GDP in Asia substantially in the medium to long term (11.7 percent when FDI reaches its steady state). - For China: - Six percentage points of the 9 percent gain in China’s real GDP is attributable to FDI liberalization. - Over 60 percent of China’s investment gains are attributable to FDI liberalization. - FDI contribution to long-term trade volumes: - Ranges from 1.2 percentage points (advanced Asia economies) to 4.6 percentage points (emerging Asia). - Consumption and investment: - FDI contributes a greater proportion of gains to investment than to consumption. - All regions gain around 2 percentage points of consumption from lower trade barriers in both the short and long term. ### Scenario and decomposition results - Scenario ordering and impact: - Real GDP gains increase with each progressive enlargement of economies liberalizing. - Opening to all economies (even without reciprocation, as in scenario 3) provides more benefits. - Decomposition of GDP into tariff and NTB effects (Figure 32) — percent deviation relative to baseline: - Charts show contributions from: - Lower intermediate goods tariffs - Add final goods tariffs - Add lower services NTBs - Total - The relative contributions vary by region: Asia, China, Advanced Asia, Emerging Asia (detailed charted percent differences available in the source). ### Policy implications and recommendations - Priorities for maximizing benefits from trade liberalization in Asia: - Focus more on service sector liberalization and the easing of FDI restrictions; these would have the greatest impact on Asia. - Broaden and deepen existing or proposed trading arrangements to emphasize services and investment liberalization. - Support multilateral liberalization while recognizing plurilaterals as an alternative route. - Complementary domestic policies to manage transition costs: - Active labor market policies, social protection, and complementary policies in education, housing, and credit to smooth labor reallocation and limit resistance from vested interests. - Targeted policies to redistribute economy-level gains by focusing on adversely affected sectors. - Trade cost reduction measures beyond tariff/NTB cuts: - Improve regional connectivity and logistics through productive infrastructure investments. - Tackle behind-the-border barriers such as customs procedures, export taxes, countervailing measures, and technical barriers to trade. - Consider steps toward a customs union or common market in Asia covering e-commerce, trade in services, and FDI to provide a regional buffer against global trade tensions. - Sectoral notes: - Low-income Asian economies would benefit most by participating in manufacturing GVCs with advanced and emerging Asian economies to move up value chains. - The digital economy increases the importance of policies on e-commerce, investment, and services trade. ### Trade patterns and trade-costs evidence (Annex highlights) - Gravity-model findings (Annex 1): - Asia over-trades in goods by a wide margin relative to most other regions; intraregional trade in goods is high, reflecting global value chains. - Asia has been under-trading in services overall and has fallen behind other regions since 2007; intraregional trade in services has also fallen behind significantly. - Tariffs (Annex 2): - Effective tariffs have come down substantially since the early 1990s in both advanced and emerging Asia, with greater dispersion in emerging Asia. - There remains scope to bring tariffs down further. - Nontariff barriers (NTBs) (Annex 2): - NTBs have been on the rise in Asia (technical barriers to trade, quotas, bailouts, state aid, trade defense measures, preferences for local industries), though increases have slowed recently. - Sources examined include WTO, World Bank Temporary Trade Barriers database, Global Trade Alert, and UNCTAD TRAINS. - World Bank data suggest the percentage of products covered by measures is lower than in the United States and Europe (as computed by Bown and Reynolds (2014)); Global Trade Alert shows more discriminatory measures than liberalizing measures historically. - Free Trade Agreements (Annex 2): - FTAs have grown in number; recent agreements extend beyond goods to include customs and trade facilitation, services, and data protection. - ASEAN FTA lowered intraregional goods tariffs on most products to 0 to 5 percent, excluding some sensitive products. - WTO Trade Facilitation Agreement and agreements like the Japan-EU Economic Partnership Agreement address trade remedies, trade in services, customs and trade facilitation, and data protection. *Source: IMF staff estimates.* ### 1. Asia Nontariff Measures ### 1. Asia Nontariff Measures ### Trends in Nontariff Measures and Trade Barriers - Figure indicators show time series of measures from 1995 through 2017 for Asia, the United States, Latin America, Europe, and global measures (Global Trade Alert, WTO, UNCTAD TRAINS). - Temporary trade barriers: bilateral and plurilateral measures tracked from 1975 through 2017, with bilateral and plurilateral percent-of-products series reported. - Number of measures and net measures are charted for 2009–2017 showing categories: Liberalizing, Discriminatory, Net measures (values span negative to positive, axis tickmarks: -500 to 200). - Trade barrier initiations series charted for 1990–2012 across Asia, Europe, Latin America, United States (axis tickmarks: 0 to 8 measures). ### Free Trade Agreements and Trade Facilitation - Asia Free Trade Agreements by scope (cumulative number) lists economies (Singapore, Korea, Australia, China, Thailand, New Zealand, Japan, Malaysia, India, Indonesia, Philippines, Brunei Darussalam, Sri Lanka, Lao P.D.R., Vanuatu, Viet Nam, Maldives, Tonga, Cambodia, Fiji, Bangladesh, Nepal, Myanmar, Mongolia) and categories: Paperless Trade Related, Formalities, Institutional Arrangement and Cooperation, Transparency. - Trade Facilitation and Paperless Trade Implementation Score reported (Score, best=100): a value shown as "33" in the figure area. ### Transportation and Logistics Costs of Trade - Average time to trade exports for emerging market economies has decreased by seven days. - Cost to export has declined by about 30 percent for emerging market economies; comparable statistics for advanced economies were broadly flat. - Liner Shipping Connectivity Index (LSCI): between 2004 and 2017, the LSCI for emerging market economies almost doubled, while it increased by about a third for advanced economies. - Asia air freight cost (export to the United States, 2010=100) series show growth in air freight costs despite lower oil prices; oil price axis referenced (US dollars). - Cost to Export (US dollars, simple average) chart covers 2006–2017 for Advanced Asia, Emerging Asia, ASEAN with vertical axis range 800 to 2,400. - Time to Export (Number of days, weighted average) chart covers 2004–2017 for Advanced Asia and Emerging and Developing Asia with vertical axis tickmarks 10 to 45 days. - Liner Shipping Connectivity (Index Increasing from 0=100, simple average) chart covers 2000–2017 with axis tickmarks 0 to 200. ### Key empirical points from figures and text - Emerging market economies: time to export decreased by seven days; cost to export declined by about 30 percent. - LSCI: emerging market economies almost doubled LSCI between 2004 and 2017; advanced economies’ LSCI increased by about a third. - Air freight costs from Asia to the United States continue to grow despite recent lower oil prices. ### Implication highlighted - Despite improvements in cost and connectivity, substantial room remains to improve overall time to export for emerging market economies by advancing paperless trade-related implementation and facilitation. --- ### Economic Models Used - Two core models: Ricardian Trade Model (RTM) and Global Integrated Monetary and Fiscal Model (GIMF). - Both models are augmented by a productivity channel linking changes in FDI restrictiveness, tariffs, and NTBs with changes in multifactor productivity. ### Global Integrated Monetary and Fiscal Model (GIMF) - Used to consider dynamic impact of trade tensions and liberalization during the transition and broader macroeconomic effects. - Model structure: - Each economy close to identical with different steady-state ratios and behavioral parameters based on a stylized data set consistent with 2015 and 2016. - Consumption dynamics: saving households and liquidity-constrained households; saving households follow overlapping generations framework (Blanchard (1985), Weil (1987), Yaari (1965)); liquidity-constrained households consume all income each period. - Services sector included: services produced from tradable and nontradable goods; priced as input for consumption domestically or exported; services demand relatively inelastic at 0.9. - Private investment uses Bernanke-Gertler-Gilchrist (1999) financial accelerator; investment accumulates to private capital stock for tradable and nontradable firms. - Nominal side: implicit Phillips’ curves and inflation-forecast-based interest rate reaction function. - Detailed government sector reproducing simplified fiscal accounts. - Trade tracked bilaterally among all regions, with flows for noncommodity goods and services and commodities; noncommodities broken into final goods (consumption and investment), consumption services, and intermediate goods; differential tariffs and NTBs can be applied to these categories. - Core GIMF economies for exercises: China, India, Japan, Korea, advanced Asia (Australia, Hong Kong SAR, Macau SAR, New Zealand, Singapore, Taiwan Province of China), emerging Asia (all other Asian and Pacific economies), the United States, rest of world bloc. - Three satellite GIMF versions use core outcomes to compute six additional economies individually: Australia, Indonesia, Malaysia, New Zealand, the Philippines, Thailand. ### Ricardian Trade Model (RTM) - Multisector, multi-economy computable general equilibrium model for long-term sectoral impact of trade policy scenarios including tariffs, NTBs (tariff-equivalents), and productivity changes. - Builds on Eaton and Kortum (2002) and Caliendo and Parro (2015); solved in differences following Dekle, Eaton, and Kortum (2007). - Calibrated to 2014 World Input-Output Dataset (Timmer and others 2015), featuring 35 sectors and 44 economies with rich input-output links and bilateral tariffs from UNCTAD TRAINS. - Asian economies included: Australia, China, India, Indonesia, Japan, Korea, Taiwan Province of China. - Sectors defined according to ISIC revision 4. Table A3.1 lists 35 sectors (1 Agriculture, forestry and logging, fishing and aquaculture; 2 Mining and quarrying; ... 35 Other service activities). --- ### Modeling Productivity and the Role of Restrictive FDI Regimes - Productivity channel calibrated using Ahn and others (2016). - Restrictive FDI regimes can impede trade liberalization and productivity gains; channel mapped to RTM and aggregated for GIMF. - Productivity channel intuition: - Higher tariffs can increase input costs or force substitution to less desirable inputs, lowering sector productivity. - Higher tariffs can induce misallocation toward sectors without comparative advantage, reducing productivity. - If an economy has a restrictive FDI regime, productivity declines due to tariffs will be less pronounced. - Productivity change specification (as presented): - Given changes in trade costs, the TFP (A_s^n) in sector s and economy n changes according to: () () () () ,, ˆ ˆˆ soutoutssoutputininssinput nnnsnns Afdifdifdifdi α βτ αβτ= + −++ − , where ,ˆ input_ns τ are tariff changes in the inputs to sector s in economy n, and ,ˆ output_nk τ are tariff changes in sector s itself. - Coefficients used (from Ahn and others (2016)): - α_out = -1.09872 - β_out = -0.05205 - α_in = -0.05205 - β_in = -0.39589 - Sensitivity to either output or input tariffs is capped at zero to preserve that higher tariff costs hurt productivity; specification could be overturned for very high levels of FDI restrictiveness when implementing the equation. --- ### Assumptions on Trade Barriers and Policy Instruments - Tariff data: - Effective tariff rates for trade tensions scenarios based on detailed tariff data at HS six- or eight-digit levels from publicly available sources; processed into sector-level effective tariff rates for the RTM’s 35 sectors via concordance tables from World Bank WITS. - Effective tariffs related to the automobile sector computed for the GIMF using UN COMTRADE bilateral sectoral imports and exports, and an assumed 25 percent tariff levied by the United States. - Footnote mentions $200 billion in US tariffs on China’s goods and China’s retaliatory tariffs on $60 billion of US goods as part of cited source material. - Trade liberalization instruments: - Three instruments: tariffs on goods, NTBs on services (quantified as effective tariff-rate equivalents), and FDI regimes that impede productivity gains. - Goods liberalization: assume participating economies reduce all tariffs to zero in any production sector where at least one Asian economy already has a zero tariff. - Services NTBs: economies grouped as Low, Average, or High restrictiveness and make corresponding cuts of 10, 20, or 30 percent in their effective tariff-rate equivalent for NTBs. - Low NTBs (10 percent cut): Australia, Cambodia, Hong Kong SAR, Japan, Korea, Lao PDR, Mongolia, New Zealand, Singapore - High NTBs (30 percent cut): India, Indonesia, the Philippines - Average NTBs (20 percent cut): all other economies under consideration - FDI restrictiveness: quantified as the percent gap between an economy’s FDI restrictiveness index and the global average using OECD sector-level data mapped to the RTM’s 35 sectors and aggregated for the GIMF’s four sectors (intermediate, consumption, investment goods, plus services excluding travel). - Coefficients computed for: Australia, Cambodia, China, Indonesia, India, Japan, Lao PDR, Myanmar, Mongolia, New Zealand, the Philippines, Taiwan Province of China, Vietnam. Other Asian economies assumed to have FDI gaps similar to economies near their per capita income level. ### Scenario Design and Dynamics - Three liberalization scenarios defined: "China Opens", "Within Asia", and "Asia Opens" (each can have variants with and without FDI liberalization loosening). - Figure A3.1 shows dynamic percent deviations relative to baseline (trade and productivity channels only) across a timeline labeled 0 to 20 (quarters/years indexing in figure) for the three scenarios, with larger gains for successively broader liberalization: - Each scenario provides more gains than the last. - Gains grow over time, with a slowdown around five years as medium-term restrictions resolve and productivity gains begin to manifest but are not yet perceived as permanent. - Once productivity gains are perceived as permanent, gains accelerate to long-term values. ### Implementation Notes for Model Experiments - Trade liberalization enters models primarily through relative price channels and government revenue channel for tariffs. - FDI liberalization enters as an input to the productivity channel derived from Ahn and others (2016). - Trade tensions baseline and escalation scenarios use effective tariff computations and, for select exercises, assumed retaliatory and tariff-large-value references noted in source materials. *Source: IMF staff calculations and the source document content.* ### 5. Real Consumption ### 5. Real Consumption ### Scenario 3 with Foreign Direct Investment Liberalization — decomposition by channels - Figures A3.2 and A3.3 present Scenario 3 with Foreign Direct Investment (FDI) liberalization, decomposed into: Direct trade channel; Add productivity channel; Add FDI liberalization; Total. - Variables shown (percent deviation relative to baseline) include: - Real GDP (percent difference) - Real Trade Volumes (percent difference) - Real Trade Volumes for Services (percent difference) - Real Consumption (percent difference) - Real Investment (percent difference) - Real Bilateral U.S. Dollars Exchange Rate (Percent difference; positive is an appreciation) - Current Account Balance (Percent difference) - Panels decompose outcomes for geographic groups and economies: Asia; China; Advanced Asia (Australia, Hong Kong SAR, Japan, Korea, New Zealand, Singapore and Taiwan Province of China); Emerging Asia (the rest of Asia excluding China). - Figures report percent deviations at multiple ticks (examples shown in the figures): 0, 1, 2, ... up to 20 for some series; -7 to 2 for investment; -0.35 to 0 for Real GDP panels; 0 to 12 for Current Account Balance; 0 to 70 for Trade Volumes. (All numeric scales preserved as presented in the figures.) ### Key empirical findings on firm dynamism and productivity in Asia (firm-level analysis) - Asia’s productivity growth slowdown has been associated with a decline in firm dynamism, the rise of “zombie” firms, and resource misallocation. - Main quantitative stylized facts: - The overall share of young firms in the Orbis sample declined from 45 percent in 2003 to 25 percent in 2015. - In Korea, the share of young firms declined from 72 percent to 49 percent over 2003–2015. - In Japan, the share of young firms remained relatively stable at about 10 percent. - Census data for Thailand indicate a decrease in the share of young firms of 12 percentage points between 2007 and 2012. - Younger firms, on average: - Have lower TFP levels but generate higher TFP growth (result holds in five of the six economies examined). - Are generally smaller, grow faster in terms of assets, have lower leverage, and invest more in intangible capital than older firms. - Are more prevalent in sectors open to trade, where competition tends to be stronger. - “Zombie congestion” (the failure of nondynamic firms to exit) has increased in several Asian economies over the last decade; zombie firms tend to be less productive, reduce aggregate productivity, dampen productivity of healthy firms, and slow efficient capital reallocation. - The global financial crisis affected firm productivity in Asia through: - Weaker global demand (trade channel) — a temporary effect. - Tighter financial conditions — a more persistent effect. - Looser financial conditions after the global financial crisis helped explain the rise of the zombie problem and a buildup of corporate leverage, implying more firms could be affected during the next financial shock. - Other factors associated with higher productivity growth: - Firms that invest more in R&D and other forms of intangible capital. - Firms located in export-oriented sectors. - Foreign-owned firms. ### Data and scope - Data source: Orbis firm-level dataset (mostly unlisted firms, around 99 percent of sample) provided by Bureau van Dijk; firm-ownership data from Orbis provided by IMF units. - Sample focused on six Asian economies for which firm-level TFP can be estimated for a relatively large number of firms: China, Japan, Korea, Malaysia, Philippines, Thailand. - Sample period referenced in the paper: 2003–15 (and data availability as of September 18, 2018 for the background paper). ### Policy recommendations (preserved exactly as presented) - Foster firm dynamism. This would require reducing barriers to entry and facilitating resource reallocation through the exit of nonviable firms. This could be achieved by (1) lifting barriers to competition in goods markets; (2) increasing labor market flexibility; (3) improving insolvency regimes; and (4) reviewing credit guarantees and other support schemes. - Address debt overhang and avoid undue buildup of leverage by (1) facilitating judicious debt resolution and corporate restructuring plans; (2) adopting adequate institutional frameworks and implementing supportive tax measures and financial sector policies; (3) implementing appropriate microprudential and macroprudential policies to avoid the excessive buildup of leverage; and (4) reducing debt bias in tax systems. - Foster innovation and trade openness by reaping the benefits of new technologies and boosting total factor productivity (TFP) through intangible capital deepening, including R&D. Policies include fiscal incentives for intangible investment, patent policy, and competition in research grants. Economies should also continue to facilitate international trade and foreign investment to stimulate competition and knowledge transfer, which would also help boost firm dynamism. In this respect, the current trade tensions discussed in IMF (2018b) could dampen trends toward intangible capital deepening in Asia by increasing policy uncertainty for firms. *Source: IMF staff estimates and analysis as presented in apd-reo-background-papers-1018 - 5. Real Consumption.* ### 1. Share of Young Firms ### 1. Share of Young Firms ### Firm dynamism and productivity (TFP) - Younger firms generally have higher productivity growth; this holds for every economy in the sample except China, where the impact of firm age on productivity is positive but not statistically significant. - Examples of estimated age effects: - A one-year-old firm in Japan is associated with 3.5 percent higher productivity growth than a 10-year old firm. - In the Philippines, a one-year-old firm versus a 10-year-old firm is associated with 1.3 percent higher TFP growth. - Convergence: Firms that start off with lower TFP levels have higher TFP growth rates, implying convergence over time toward more productive firms. ### Other firm characteristics affecting productivity - Firm leverage: - A marginal increase in leverage is associated with higher productivity growth during normal times; the positive effect is smaller when firms start from higher leverage levels. - During crises, the impact of leverage on productivity becomes negative and persists beyond the crisis. - Investment in intangibles: - Intangible capital deepening leads to higher productivity growth. - The effectiveness of an additional unit of intangible capital (as a share of fixed assets) varies over time, across economies, and across industry types. - Intangibility effects are higher in companies with higher intangible capital; higher in Japan than in Korea, though the difference has declined since the global financial crisis. - Intangible capital appears to contribute significantly more to productivity in Germany, Italy, and Hungary than in Japan and Korea. - Note: Intangible assets in Orbis include R&D expenses, goodwill, and other long-term expenses; R&D data limitations lead to focus on intangible assets for Japan, Korea, Thailand, and Malaysia. - Openness to trade: - Firms in sectors with higher export shares in total output have higher productivity growth than those in lower trade-intensity sectors. - Firms more exposed to global demand may also be more affected by global shocks. - Foreign ownership: - Defined as nonresident ownership of more than 50 percent of a firm’s capital. - Foreign ownership is associated with 0.5 percent higher TFP growth. - Firm size: - Larger firms in general exhibit higher productivity growth, controlling for firm age. ### Quantitative contributions (selected figures) - Contribution of average firm age to TFP growth is estimated and differs by country; graphical results show Japan, Korea, China, Thailand, Malaysia, Philippines, Germany, Italy, Hungary. - Contribution of average intangibility to TFP growth is estimated and shown for countries with statistically significant coefficients (at the 10 percent level). - Figure-based observations in the source document report R&D (gross domestic expenditure on R&D) time series for China, Japan, Republic of Korea, Malaysia, Philippines, Thailand (percent of GDP). ### Panel regression approach - A panel regression is employed to assess links between firm characteristics and TFP growth; methodological setup described in Annex 3. - Results indicate younger firms, higher intangibility share, operation in more export-intensive sectors, foreign ownership, firm size, and initial TFP level (convergence) all matter for productivity growth. - Due to scarce firm-level exports data, openness analysis uses China, Japan, and Korea with industry-level export-to-gross-output ratios from the World Input-Output Database. ### Note on numeric precision - All numeric values, percentages, and definitions are preserved exactly as in the source. ### Zombie firms and resource misallocation - Definition used (baseline, OECD-based): a firm is a zombie in 2015 if it is 10 years old or older and has had an interest coverage ratio of less than one for three consecutive years (2013–15). Interest coverage ratio = EBIT / interest paid. - Key findings on zombies (Asia and comparisons): - Japan: Zombies constitute about 10 percent of firms and account for 8 percent of total capital stock. - Thailand: Similar magnitude to Japan in terms of zombie problem. - Korea: Total number of zombie firms is small, but the share of capital resources sunk into these firms is about 7 percent. - China: Zombie capital share recently reached 6 percent. - Europe: Capital shares of zombies appear to be higher in some European countries; comparisons should be interpreted with care due to idiosyncratic factors. - Characteristics of zombie firms (2014): - Zombies tend to be older, larger, more leveraged, and have less intangible investment than non-zombies. - Zombies are predominant in traditional sectors: petroleum products, chemicals, printing and paper, textiles, basic metals, and (in China) shipbuilding; less prevalent in knowledge-intensive sectors like computers, electronics, and machinery. - Estimated productivity differential: - Controlling for leverage, size, age, and intangibility, productivity growth is 1.4 percentage points higher among non-zombies than among zombies. - Spillovers and simulated impacts: - A higher number of zombie firms and a larger share of capital sunk into zombie firms have negative effects on healthy firms’ productivity and tangible and intangible investment. - Counterfactual simulation: if zombie shares had stayed at their 2005 level, the increase in yearly growth of tangible investment in healthy firms by 2014 would have ranged from 0.07 percent in Korea to 0.25 percentage points in China. - Productivity growth would have been up to 0.1 percent higher in 2014 under that counterfactual. - Potential gains from halving zombie capital share in a representative four-digit industry in 2014: - Cutting zombie shares by half could increase annual TFP growth by between 0.03 percent (Malaysia) and 0.09 percent (Thailand). - Reducing zombie shares could increase growth in tangible investment between 0.08 and 0.21 percentage points per year. - Benchmark median non-zombie firm in 2014: - TFP growth of 0.14 percentage points. - Tangible investment growth of -0.69 percentage points. - Intangible investment growth of zero percentage points. - Interpretation: Performance of a typical non-zombie firm would increase by between 20 to 60 percent in terms of TFP growth and 10 to 30 percent in terms of tangible investment growth if zombie shares were halved (in the representative-industry scenario). - Resource misallocation effects: - Evidence that a higher presence of zombies reduces the ability of more productive firms to increase tangible investment. - Cutting zombie shares by half allows the difference in capital growth between high- and low-productivity firms to increase (example lower bound: 0.3 percentage points for Malaysia; additional country-specific figures follow in the source). ### Policy-relevant implications (from findings) - Addressing zombie firms could yield measurable gains in aggregate productivity and investment: - Restructuring or resolving zombie firms can free up capital and resources for more productive firms and startups. - Policies that limit prolonged forbearance and encourage efficient corporate restructuring may reduce misallocation and support creative destruction. - Promoting intangible investment and openness to trade, along with facilitating foreign participation, can raise productivity growth, especially when coupled with policies that prevent resource crowding by zombies. - Monitoring firm-level leverage and reinforcing financial sector incentives for timely recognition of nonviable firms can mitigate long-term negative productivity effects from crises. *Source: IMF staff estimates and analysis (apd-reo-background-papers-1018 - 1. Share of Young Firms).* ### 0.9 percentage points (for Thailand) (Figure 13). ### apd-reo-background-papers-1018 - 0.9 percentage points (for Thailand) (Figure 13) ### Executive summary and main findings - Waning firm dynamism has been an important factor contributing to the productivity slowdown in Asia. - The share of young firms has declined and zombie firms have increased significantly in several economies, hampering efficient resource allocation. - Capital growth rates at the high-productivity firms would increase between 1.5 and 4 percentage points under reallocation simulations. - Cutting the zombie share by half in 2014 yields potential gains in TFP growth and capital reallocation for non-zombie firms (see Figures 11–13 in source). ### Firm-level drivers: young firms, zombies, and intangible investment - Young firms are generally more productive; their falling share contributed to the slowdown. - Zombie firms: - Have lower productivity growth and hamper resource allocation toward more productive firms. - Increasing zombie shares caused counterfactual losses for typical non-zombie firms (Figure 11). - Higher productivity growth at the firm level is associated with: - Investing more in intangible assets (for example, R&D). - Belonging to exporting sectors. - Being foreign-owned. - Simulation note: The simulations compare capital accumulation rates between firms at the 75th percentile and 25th percentile of the TFP distribution (high- and low-productivity firms respectively). ### Sectoral shifts and their contribution to aggregate productivity - Two ways sectors contribute to aggregate productivity: within-sector productivity growth and between-sector reallocation effects. - Decomposition compares 2003–08 (pre-GFC) with 2011–16 (post-GFC) across agriculture, industry, and services. - Main sectoral findings: - Total labor productivity growth rates in Asia slowed after the global financial crisis for all but one examined economy (the Philippines). - In China, a significant part of the slowdown between 2002–08 and 2011–16 reflected smaller gains from shifting resources from agriculture to manufacturing, because labor shifts increasingly went to services rather than manufacturing. - In other examined Asian economies (excluding Thailand due to a structural break), sectoral shifts contributed little to productivity gains; the slowdown mainly reflected a within-manufacturing productivity slowdown (e.g., Korea and Malaysia saw within-manufacturing contributions decline by more than 1.3 percentage points between 2003–08 and 2011–15). ### Global financial crisis (GFC): channels and quantified impacts - Two channels examined: weaker global demand (trade openness) and tightening financial constraints (precrisis leverage). - Role of weak global demand: - Firms in more open/export-oriented sectors experienced a larger TFP growth slowdown during the crisis years (2008–09). - Firms in sectors where 14.3 percent of gross output was exported experienced a greater slowdown by 0.3 percentage points compared with firms in sectors with virtually no precrisis exports. - The crisis-era trade-related slowdown was not persistent into 2010–15 on average. - Role of financial constraints (precrisis leverage defined as total liabilities/total assets): - A firm with leverage equal to the 2007 average (61 percent) experienced a TFP slowdown 0.4 percentage points larger than firms with no leverage during the crisis. - The impact of leverage on TFP growth persisted into 2010–15, with an impact close to 0.5 percentage points for firms at 57 percent leverage. - Precrisis leverage negatively affected investment: - Tangible investment growth during 2008–09 slowed by 2.5 percentage points more for firms with leverage at the 2007 average compared with firms with no leverage. - Precrisis leverage had a negative and statistically significant impact on intangible investment growth during 2008–09, though not statistically significant thereafter; the temporary reduction in intangible investment had persistent effects on the stock of intangible assets and thus on productivity. ### Macro-level context: corporate debt and risks - Post-GFC loose monetary and easy international credit conditions contributed to an increase in corporate debt in many Asian economies (see Figure 18). - Rising corporate debt, if left unaddressed, poses a significant risk to productivity growth when financial conditions tighten. ### Policy implications and recommended priorities - Three broad, interrelated policy areas for Asian policymakers: 1. Foster firm dynamism by facilitating entry and exit: - Lift barriers to competition in goods markets and promote labor market flexibility to facilitate reallocation. - Reform product and service sector regulations to reduce barriers to entrepreneurship and downstream effects that impede manufacturing productivity. 2. Address debt overhang and excessive leverage: - Improve foreclosure and insolvency frameworks, facilitate judicious debt resolution and corporate restructuring plans, limit bank forbearance, and resolve nonperforming loans. - Reassess policies that may prevent firm exit (including some SME support programs) where they perpetuate low-productivity firms. - Consider well-targeted tax policies to shift capital structures from debt toward equity to reduce leverage and boost investment in tangible and intangible capital. - Adopt adequate institutional frameworks and micro- and macroprudential policies to avoid excessive leverage buildup. 3. Foster innovation and trade openness: - Encourage intangible investment (including R&D) via fiscal incentives where appropriate, broaden scope of existing incentives, and ensure intellectual property regimes balance innovation incentives and competition. - Consider competition in research grants and assess direct subsidies for intangible investment carefully, given risks that R&D subsidies to incumbents can encourage survival of low-productivity firms. - Continue opening to international trade and foreign investment to stimulate competition, knowledge transfer, and higher productivity growth. ### Quantified examples and reference figures (selected) - Capital growth at high-productivity firms: increase between 1.5 and 4 percentage points in simulations. - Trade openness example: sector with 14.3 percent of gross output exported saw a 0.3 percentage points larger TFP slowdown during 2008–09 relative to sectors with virtually no exports. - Precrisis leverage benchmarks and impacts: - 2007 average leverage: 61 percent → 0.4 percentage points larger TFP slowdown during crisis. - 57 percent leverage → impact close to 0.5 percentage points on TFP growth in 2010–15. - Tangible investment slowdown for firms at 2007 average leverage: 2.5 percentage points larger during 2008–09. *Source: IMF staff analysis as presented in apd-reo-background-papers-1018 - 0.9 percentage points (for Thailand) (Figure 13).* ### Annex 1. Theoretical Framework for Firm Dynamism ### Annex 1. Theoretical Framework for Firm Dynamism ### Schumpeterian/Hopenhayn model intuition - Firms face idiosyncratic productivity shocks as a source of uncertainty. - Entry cost E per unit of capital creates entrant profit π_e = a_e k − r k − E k; entry occurs if a_e > r + E, creating a wedge between the lower limit of productivity to enter and the rental rate of capital. - Exit cost X per unit of capital creates incumbent exit profit π_ix = −X k if exit, and continuation profit π_ip = a_i k − r k if operate; incumbents continue if a_i > r − X. - High entry costs reduce entry and protect incumbents (especially large firms), lowering competitive pressure and dampening innovation; combination of entry and exit costs can give rise to zombie congestion and lower aggregate productivity. ### Formal production specification - Firm production: y_n = a_n k. - Entrant profit: π_e = a_e k − r k − E k; entry if a_e > r + E. - Incumbent continuation condition: a_i > r − X. --- ### Annex 2. Details on the Firm-Level Dataset ### Data source and sample - Firm-level data: Orbis (around 99 percent unlisted firms), provided by Bureau van Dijk. - Sample economies: China, Japan, Korea, Malaysia, the Philippines, and Thailand. - Time span: 2003–15. - Sector coverage: nonfarm, nonfinancial business sector corresponding to two-digit industry codes 5-82 in NACE Revision 2 (including manufacturing and selected services). ### Data cleaning - Two main steps: (1) cleaning reporting mistakes (eliminate observations with negative real variables, e.g., negative total assets); (2) quality checks on consistency of balance sheet variables. - Cleaning follows procedures in Gal (2013), Kalemli-Özcan and others (2015), and Gopinath and others (2017) as implemented by Diez, Fan, and Villegas-Sanchez (forthcoming). ### Representativeness - Orbis broadly comparable to census data in sectoral distribution for manufacturing, wholesale, and construction (figures shown for Korea 2015, Japan 2012, China 2014, Philippines 2010). --- ### Annex 3. Baseline Panel Regression and Firm-Level Findings ### Baseline regression specification - Equation (A3.1): ΔTFP_i c s t = β x_i c s t + γ_i + δ_c s t + ε_i c s t. - Dependent variable: annual growth in TFP for firm i. - x includes: intangibility of fixed assets, logarithm of age, log total assets, lagged TFP level, leverage (linear and square), and other firm-level controls. - Controls: firm fixed effects γ_i; country-sector-year fixed effects δ_c s t (sector = four-digit NACE Revision 2); standard errors clustered at firm level. ### Firm-level variable definitions (selected) - TFP: Estimated based on the method proposed by Gandhi, Navarro, and Rivers (2017). - Age: current year − date of incorporation + 1 (log used). - Interest coverage ratio (ICR): EBIT / interest paid. - Leverage: total liabilities / total assets. - Intangibility: intangible fixed assets / total fixed assets. - Size: total assets (USD, log). - Intangible investment: net change in intangible fixed assets normalized by lagged total assets. - Tangible investment: net change in tangible fixed assets normalized by lagged total assets. - Foreign ownership: dummy if > 50 percent owned by foreigners. ### Key regression estimates (Table A3.2, Panel A: Firm-level drivers of TFP) - TFP t-1: −0.621*** (0.003) - Leverage t-1: 0.062*** (0.003) - Leverage^2 t-1: −0.007** (0.003) - Intangibility t-1: 0.015*** (0.003) - Total assets: 0.060*** (0.001) - Age: −0.023*** (0.001) - Country-Industry-Year F.E.: Yes - Firm F.E.: Yes - Observations: 3 291 433 - R square: 0.452 - Note: ***, **, and * indicate 1, 5, and 10 percent. ### Foreign ownership and trade intensity (Panels B and C) - Panel B: Foreign ownership - Foreign: 0.005*** (0.001) - Country-Industry-Year F.E.: Yes - Firm controls: Yes - Observations: 488 744 - R square: 0.105 - Panel C: Trade intensity - Trade intensity t-1: 0.033*** (0.005) - Country-Year F.E.: Yes - Country-Industry F.E.: Yes - Industry-Year F.E.: Yes - Firm controls: Yes - Firm F.E.: Yes - Observations: 2 704 922 - R square: 0.473 ### Intangibility subsample results (Table A3.5) - Sample with Min(Intangibility) > 1 percent: - Intangibility t-1: 0.042*** (0.007) - Country-Industry-Year F.E.: Yes - Firm F.E.: Yes - Observations: 318 223 - R square: 0.485 - Sample with Min(Intangibility) > 10 percent: - Intangibility t-1: 0.077*** (0.016) - Country-Industry-Year F.E.: Yes - Firm F.E.: Yes - Observations: 48 044 - R square: 0.304 - Figure A3.1 notes: - Coefficient × mean intangibility: 13.1 percent in the >1% sample and 43.6 percent in the >10% sample. - Interquartile variation in intangibility is 1 percentage point. ### Additional illustrative impacts (from figures) - Impact on TFP growth from interquartile increase in RHS variables (in percent): - Leverage: 2.2 - Total assets: 12.5 - Age: −3.0 - Impact on TFP growth from foreign ownership and trade intensity (in percent): - (Assumes variation in trade intensity from 25th percentile 0.2 percent to 75th percentile 17 percent.) --- ### Annex 4. Cross-Country Comparison of Firm-Level Productivity Drivers ### Main cross-country findings (based on balanced country panels; Table A4.1 and Figure A4.1) - Leverage: - Linear leverage term positive and quadratic leverage term negative in all economies, consistent with increasing net benefits at low leverage and decreasing benefits at high leverage. - Linear term statistically significant at 1 percent for all Asian economies. - Quadratic term statistically significant at 1 percent for Japan, Korea, and China; significance lower for Thailand and Malaysia. - Contribution of leverage to productivity growth varies across economies; larger for Malaysia and Philippines, very small for China. - Age: - In most economies, age coefficient negative: younger firms tend to have higher productivity growth. - China is an exception: age coefficient positive, implying younger firms show lower productivity growth (possibly due to many entrants exiting quickly). - Intangibility: - Heterogeneous effects across economies. - Only Japan and Korea show positive and statistically significant impact of intangible capital; impact smaller for Korea. - Some European economies (Germany, Italy, Hungary) show larger positive impact of intangibility than Asian economies. - Recent subsample evidence (unreported) suggests intangibility impact has become more prominent recently in Japan and Korea. - Other controls: - Negative coefficients on lagged productivity confirm productivity convergence. - Positive coefficients on firm size indicate larger firms have higher productivity growth. - Each factor’s contribution to productivity growth does not change a lot over time within firms. ### Selected country-specific coefficient estimates (Table A4.1) - TFP Level t-1: - Japan: −0.5955*** (0.0008) - Korea: −0.6651*** (0.0014) - China: −0.8553*** (0.0027) - Thailand: −0.5539*** (0.0019) - Malaysia: −0.6775*** (0.0145) - Philippines: −0.7956*** (0.0301) - Leverage t-1: - Japan: 0.1376*** (0.0032) - Korea: 0.0572*** (0.0056) - China: 0.0184*** (0.0041) - Thailand: 0.0698*** (0.0077) - Malaysia: 0.2935*** (0.0917) - Philippines: 0.4018*** (0.1029) - Leverage t-1 × Leverage t-1 (quadratic): - Japan: −0.0419*** (0.0027) - Korea: −0.0152*** (0.0053) - China: −0.0136*** (0.0039) - Thailand: −0.0146* (0.0080) - Malaysia: −0.1575 (0.0916) - Philippines: −0.2983*** (0.0951) - Size t: - Japan: 0.0844*** (0.0004) - Korea: 0.0497*** (0.0007) - China: 0.0064*** (0.0005) - Thailand: 0.0794*** (0.0011) - Malaysia: 0.1348*** (0.0114) - Philippines: 0.0267** (0.0111) - Age t: - Japan: −0.0208*** (0.0012) - Korea: −0.0297*** (0.0017) - China: 0.0420*** (0.0024) - Thailand: −0.0115*** (0.0034) - Malaysia: −0.0044 (0.0353) - Philippines: −0.0079 (0.0340) - Intangibility t: - Japan: −0.0398*** (0.0022) - Korea: 0.0143*** (0.0032) - China: −0.0007 (0.0025) - Thailand: 0.0072 (0.0100) - Malaysia: −0.0188 (0.0352) - Philippines: 0.0396 (0.0794) - Time periods and sample sizes: - Japan: 2002-2015, Observations 1,406,707, R-squared 0.446 - Korea: 2003-2015, Observations 569,356, R-squared 0.456 - China: 2005-2014, Observations 264,390, R-squared 0.620 - Thailand: 2003-2015, Observations 294,919, R-squared 0.376 - Malaysia: 2006-2014, Observations 5,976, R-squared 0.536 - Philippines: 2009-2015, Observations 1,789, R-squared 0.677 - Significance: * significant at 10%, ** significant at 5%, *** significant at 1%. *Source: apd-reo-background-papers-1018 - Annex 1. Theoretical Framework for Firm Dynamism* ### 2. Impact of Intangibility on TFP Growth across Countries ### 2. Impact of Intangibility on TFP Growth across Countries ### Cross-country estimated impacts - Estimated coefficients multiplied by mean value based on balanced country panel data. - Countries with statistically significant coefficients (at the 10 percent level) are shown for comparisons Before 2008 and After 2009 (countries listed: Japan, Korea, China, Thailand, Malaysia, Philippines, Germany, Italy, Hungary, Colombia, Full sample). (No numeric coefficients for countries provided in source extract.) ### Firm-size effects on TFP - The estimated coefficients multiplied by mean value based on balanced country panel data are reported for firm size effects (chart scale shows up to 30 percent). (No numeric coefficients provided in source extract.) --- ### Annex 5. Zombie Firms ### Definition and data construction - A firm is defined as a zombie if it is 10 years old or older and has had an interest coverage ratio (ICR) less than one for three consecutive years. - ICR is estimated using financial expense data (sum of interest paid and write-off of financial assets) in Orbis. Where both interest paid and financial expense are available, the average ratio of interest paid to financial expense is used to approximate interest payments for a wider sample. - Analysis focuses on manufacturing firms (NACE Revision 2, four-digit sectors 10-33). ### Empirical specification - Two baseline regressions estimated (equations A5.1 and A5.2): - Equation A5.1: Y_i s c t = β1 nonzombie_i s c t + β2 nonzombie_i s c t * n_c + β3 TFPrfcontrols_{i s c t-1} + δ_{i s c t} + ε_{i s c t} - Y refers to a measure of activity (productivity growth, tangible and intangible investment in percent of total assets) in firm i, industry s, country c, time t. - n_c is either the number of zombie firms or the share of industry capital sunk in zombie firms. - TFPrfcontrols_{i s c t-1} are lagged controls: leverage, share of intangible assets in total assets, firm size (total assets), firm age (years since incorporation). - Equation A5.2: Kgrowth_{i s c t} = γ1 TFP_{i s c t-1} + γ2 TFP_{i s c t-1} * n_c + γ3 TFPrfcontrols_{i s c t-1} + δ_{i s c t} + ε_{i s c t} - Kgrowth is change in real tangible or intangible capital stock. - Interacted country-industry-year fixed effects included. Robust standard errors clustered at country-industry-year level. ### Expected signs and interpretation - β2 expected to be negative if zombie congestion reduces non-zombie firms’ investment and productivity growth. - β1 expected to be positive (non-zombie firms more productive), but could be negative if zombie firms receive increasingly large subsidies. - γ1 expected to be positive (higher productivity attracts resources and growth). - γ2 expected to be negative if zombie firms distort capital reallocation efficiency. ### Key regression results (selected reported coefficients) - Panel C (Capital allocation and capital share of zombie firms) - TFP_{t-1}: 0.264*** (standard error 0.012) in column (1) Tangible investment; 0.001 (standard error 0.001) in column (2) Intangible investment. - TFP_{t-1} * Zombie share: -0.242*** (standard error 0.022) in column (1) Tangible investment; -0.001 (standard error 0.003) in column (2) Intangible investment. - Observations: 954 723 (column 1), 845 694 (column 2). - R square: 0.424 (column 1), 0.431 (column 2). - Panel A (Capital owned by zombie firms) - Nonzombie: 0.014*** (standard error 0.001) for TFP growth; 0.013*** (standard error 0.001) for Tangible investment; 0.001** (standard error 0.000) for Intangible investment. - Nonzombie * Zombie share: -0.016*** (standard error 0.006) for TFP growth; -0.039*** (standard error 0.009) for Tangible investment; -0.003** (standard error 0.001) for Intangible investment. - Observations: 1 104 651 (TFP growth), 1 099 617 (Tangible), 979 519 (Intangible). - R square: 0.085 (TFP growth), 0.014 (Tangible), 0.084 (Intangible). - Panel B (Number of zombie firms) - Nonzombie: 0.014*** (standard error 0.001) for TFP growth; 0.034*** (standard error 0.002) for Tangible investment; 0.001 (standard error 0.001) for Intangible investment. - Nonzombie * Zombie share: -0.014* (standard error 0.008) for TFP growth; -0.258*** (standard error 0.015) for Tangible investment; -0.004 (standard error 0.003) for Intangible investment. - Observations: 1 104 651 (TFP growth), 1 099 617 (Tangible), 979 519 (Intangible). - R square: 0.085 (TFP growth), 0.014 (Tangible), 0.084 (Intangible). ### Distributional evidence (Figure A5.1 highlights) - Economy-level distributions of zombie firms by sector reported for Japan, Korea, China, Thailand, Malaysia, and others: - Charts report Number of firms and Share in capital by sector (Food; Beverages; Textiles; Apparel; Leather; Wood; Paper; Printing; Refined petroleum products; Chemicals; Basic pharmaceuticals; Rubber and plastic products; Mineral products; Basic metals; Metal products; Computers and electronics; Electrical products; Machinery and equipment; Motor vehicles; Other transport equipment; Furniture; Other manufacturing; Repairs and installation). - Japan-specific charts include age ratio and size ratio series (Age ratio and Size ratio scales shown in panels 1–2). - Korea, China, Thailand, Malaysia panels likewise show Number and Share in Capital and Age and Size ratios for zombies by sector. - Sources: Orbis; and IMF staff calculations. --- ### Annex 6. Impact of the Global Financial Crisis: Differences-in-Differences Model ### Empirical approach - Differences-in-differences model comparing TFP growth differences between firms with large versus low precrisis balance sheet vulnerabilities, before and after 2008 credit tightening. - Baseline regression (A6.1): dependent variable is difference in average annual firm TFP growth between postcrisis and precrisis periods. - Precrisis period defined as 2003–07. - Postcrisis periods used: - Baseline: 2008–15 - Crisis window: 2008–09 - Persistence window: 2010–15 - Variable of interest: precrisis average leverage for firm i (proxy for precrisis balance sheet vulnerabilities). - Expectation: β1 negative (firms with higher precrisis leverage face lower postcrisis TFP growth). - Controls: precrisis average firm-level characteristics, country-sector fixed effects; standard errors clustered at country-sector level. ### Main results on leverage and TFP (Table A6.1) - Pre-crisis Leverage: - (1) 2008-2015: -0.010*** (standard error 0.001) - (2) 2008-2009: -0.007*** (standard error 0.001) - (3) 2010-2015: -0.009*** (standard error 0.001) - Country-Industry F.E.: Yes (all columns) - Firm controls: Yes (all columns) - Observations: 333 077 (column 1), 302 525 (column 2), 246 774 (column 3) - R square: 0.028 (col 1), 0.03 (col 2), 0.03 (col 3) - Figure A6.1 note: Impact of Leverage on Post-GFC TFP Growth (In percent, 2008–15). Assumes interquartile increase in leverage ratio, from 40 to 77 percent. ### Leverage and investment (Panel B; Table A6.2 summary) - Dependent variable: change in average annual investment in postcrisis vs precrisis periods (tangible and intangible separately). Precrisis average investment controlled for. - Selected coefficients (panel B): - Pre-crisis Leverage (Tangible investment, 2008-2009): -0.047*** (standard error 0.007) - Pre-crisis Leverage (Tangible investment, 2010-2015): -0.019*** (standard error 0.003) - Pre-crisis Leverage (Intangible investment, 2008-2009): -0.001*** (standard error 0.000) - Pre-crisis Leverage (Intangible investment, 2010-2015): -0.001 (standard error 0.000) - Country-Industry F.E.: Yes for tangible; Yes for intangible where applicable. - Observations and R squares: - Tangible 2008-2009: Observations 315 864; R square 0.453 - Tangible 2010-2015: Observations 243 823; R square 0.766 - Intangible 2008-2009: Observations 240 364; R square 0.768 - Intangible 2010-2015: Observations 194 139; R square 0.649 - Figure A6.2 note: Impact of Leverage on Post-GFC Investment Growth (In percent, 2008–15). Assumes interquartile increase in leverage ratio, from 40 to 77 percent. ### Trade openness channel (Panel C; Table A6.3 summary) - Investigates precrisis average trade openness (trade intensity) as explanatory variable for change in TFP growth. - Pre-crisis Trade Intensity coefficients: - (1) 2008-2009: -0.022*** (standard error 0.007) - (2) 2010-2015: 0.001 (standard error 0.007) - Country and Industry fixed effects used separately in this specification (instead of country-sector interaction fixed effects). Standard errors clustered at sector level. - Observations: 255 827 (2008-2009), 202 133 (2010-2015) - R square: 0.018 (2008-2009), 0.025 (2010-2015) - Figure A6.3 note: Impact of Trade Intensity on 2008–09 TFP Growth (In percent). Assumes interquartile increase in export/output ratio, from 0.7 to 20 percent. --- *Source: apd-reo-background-papers-1018 - 2. Impact of Intangibility on TFP Growth across Countries; apd-reo-background-papers-1018 - Annexes 5 and 6 (Orbis; and IMF staff calculations).* ### References ### apd-reo-background-papers-1018 - References ### Innovation, productivity, and firm dynamics - Acemoglu, Daron, Ufuk Akcigit, Harun Alp, Nicholas Bloom, and William R. Kerr. 2017. “Innovation, Reallocation, and Growth.” CEPR Discussion Paper No. 12449, Centre for Economic Policy Research, London. - Alcalá, F., and A. Ciccone. 2004. “Trade and Productivity.” The Quarterly Journal of Economics 119(2): 613–46. - Andrews, Dan, Chiara Criscuolo, and Peter N. Gal. 2015. “Frontier Firms, Technology Diffusion and Public Policy: Micro Evidence from OECD Countries.” In OECD: The Future of Productivity. Main Background Papers. Paris: Organization for Economic Cooperation and Development. - Andrews, Dan, Chiara Criscuolo, and Peter N. Gal. 2016. “The Best versus the Rest: The Global Productivity Slowdown, Divergence Across Firms and the Role of Public Policy.” OECD Productivity Working Paper No. 5, OECD Publishing, Paris. - Andrews, Dan, Müge Adalet McGowan, and Valentine Millot. 2017. “Confronting the Zombies: Policies for Productivity Revival.” OECD Economic Policy Paper No. 21, OECD Publishing, Paris. - Brandt, Loren, Johannes Van Biesebroeck, and Yifan Zhang. 2012. “Creative Accounting or Creative Destruction? Firm-level Productivity Growth in Chinese Manufacturing.” Journal of Development Economics 97(2): 339–51. - Decker, Ryan A., John Haltiwanger, Ron S. Jarmin, and Javier Miranda. 2016. “Declining Business Dynamism: Implications for Productivity?” Hutchins Center on Fiscal and Monetary Policy at Brookings Working Paper No. 23, Brookings Institution, Washington, DC. - De Loecker, Jan. 2013. “Detecting Learning by Exporting.” American Economic Journal: Microeconomics 5(3): 1–21. - Foster, Lucia, John C. Haltiwanger, and C. J. Krizan. 2001. "Aggregate Productivity Growth: Lessons from Microeconomic Evidence," NBER Chapters, in: New Developments in Productivity Analysis, pages 303-372. - Gal, Peter N. 2013. "Measuring Total Factor Productivity at the Firm Level using OECD-ORBIS," OECD Economics Department Working Papers 1049, OECD Publishing. - Hallward-Driemeier, Mary, Giuseppe Iarossi, and Kenneth L. Sokoloff. 2002. “Exports and Manufacturing Productivity in East Asia: A Comparative Analysis with Firm-level Data.” NBER Working Paper No. 8894, National Bureau of Economic Research, Cambridge, MA. - Hopenhayn, Hugo A. 1992. “Entry, Exit, and Firm Dynamics in Long Run Equilibrium.” Econometrica 60(5): 1127–150. - Timmer, Marcel P. and Gaaitzen J. de Vries. 2009. “Structural change and growth accelerations in Asia and Latin America: a new sectoral data set.” Cliometrica 3: 165-190. ### Zombie firms, misallocation, and credit frictions - Ahearne, Alan G., and Naoki Shinada. 2005. “Zombie Firms and Economic Stagnation in Japan.” International Economics and Economic Policy 2(4): 363–81. - Caballero, Ricardo J., Takeo Hoshi, and Anil K. Kashyap. 2008. “Zombie Lending and Depressed Restructuring in Japan.” American Economic Review 98(5): 1943–977. - Chen, Binkai, and Guangjun Shen. 2017. “Zombie Firms and Over-capacity in Chinese Manufacturing.” China Economic Review 44 (July): 327–42. - Fukuda, Shin-ichi, and Jun-ichi Nakamura. 2011. “Why Did ‘Zombie’ Firms Recover in Japan?” The World Economy 34(7): 1124–137. - Kwon, Hyeog, Ug, Futoshi Narita, and Machiko Narita. 2015. “Resource Reallocation and Zombie Lending in Japan in the 1990s.” Review of Economic Dynamics 18: 709–32. - Lam, W. Raphael, Alfred Schipke, Yuyan Tan, and Zhibo Tan. 2017. “Resolving China’s Zombies: Tackling Debt and Raising Productivity.” IMF Working Paper 17/266, International Monetary Fund, Washington, DC. - McGowan, Adalet Müge Adalet, Dan Berlingieri and Valentine Millot. 2017. "The Walking Dead?: Zombie Firms and Productivity Performance in OECD Countries", OECD Economics Department Working Paper No. 1372, OECD Publishing, Paris. - Peek, Joe, and Eric S. Rosengren. 2005. “Unnatural Selection: Perverse Incentives and the Misallocation of Credit in Japan.” American Economic Review 95(4): 1144–1 66. - Tan, Yuyan, Yiping Huang, and Wing Thye Woo. 2016. “Zombie Firms and the Crowding-Out of Private Investment in China.” Asian Economic Papers 15(3). ### Automation, robots, and the future of work - Acemoglu and Restrepo (approach cited in the accompanying paper) — methodology referenced in analysis of robots and employment effects. - B. References to robot-related production and density in Asia are underpinned by sectoral and country studies cited across the references list (e.g., industry-focused and cross-country productivity and technology diffusion studies). - Empirical foundations for automation impacts draw on firm- and sector-level analyses cited above (e.g., Brandt et al. 2012; Decker et al. 2016; Foster et al. 2001). ### E-commerce, Fintech, and digitalization of services - Arnold, Jens Matthias, Beata Javorcik, Molly Lipscomb, and Aaditya Mattoo. 2014. “Services Reform and Manufacturing Performance: Evidence from India.” The Economic Journal 126(590): 1–39. - Di Mauro, Filippo, Duy Hoang, Andy Feng, Swee Joo Ong, Jeannette Pang. 2018. “Productivity was not that Sluggish in Developing Asia, Afterall. A Firm Level Perspective Using a Novel Dataset.” Productivity Research Network, National University of Singapore mimeo. - Doms, Mark, and J. Bradford Jensen. 1998. “Comparing Wages, Skills, and Productivity between Domestically and Foreign-Owned Manufacturing Establishments in the United States.” In Geography and Ownership as Bases for Economic Accounting, edited by Baldwin, Robert E., Robert E. Lipsey, and J. David Richardson. Chicago: University of Chicago Press. - Dougherty, Sean, Veronica Frisancho, and Kala Krishna. 2013. “State-level Labor Reform and Firm-level Productivity in India.” India Policy Forum 2013–14. - Gopinath, Gita, Şebnem Kalemli-Özcan, Loukas Karabarbounis, and Carolina Villegas-Sanchez. 2017. "Capital Allocation and Productivity in South Europe," The Quarterly Journal of Economics, Oxford University Press, vol. 132(4), pages 1915-1967. - Rao, Nirupama. 2016. “Do Tax Credits Stimulate R&D Spending? The Effect of the R&D Tax Credit in Its First Decade.” Journal of Public Economics 140(C): 1–12. ### Public finance, tax systems, and digitalization - International Monetary Fund (IMF). 2017. “Upgrading the Tax System to Boost Productivity,” IMF Fiscal Monitor Chapter 2, April 2017. - IMF. 2018a. “Asia as the Cycle Matures.” Asia and Pacific Department Regional Economic Outlook Background Paper No. 1. Washington, DC, October. - IMF. 2018b. “Regional Economic Outlook Asia and Pacific Background Paper No. 2—The Evolving Role of Trade in Asia: Opening a New Chapter” (Washington, DC). - IMF. 2018c. “Regional Economic Outlook Asia and Pacific Background Paper No. 4—The Digital Revolution in Asia: Disruptor or New Growth Engine (or Both)?” (Washington, DC). - Koske, Isabell, Isabelle Wanner, Rosamaria Bitetti, and Omar Barbiero. 2015. “The 2013 Update of the OECD Product Market Regulation Indicators: Policy Insights for OECD and non-OECD Countries.” OECD Economics Department Working Paper No. 1200, OECD Publishing, Paris. - Lanau, Sergi, and Petia Topalova. 2016. “The Impact of Product Market Reforms on Firm Productivity in Italy.” IMF Working Paper No. 16/119, International Monetary Fund, Washington, DC. ### Methods, datasets, and construction of firm-level data - Kalemli-Ozcan, Sebnem, Bent Sorensen, Carolina Villegas-Sanchez, Vadym Volosovych, and Sevcan Yesiltas. 2015. "How to Construct Nationally Representative Firm Level data from the ORBIS Global Database," NBER Working Papers 21558, National Bureau of Economic Research, Inc. - Kalemli-Ozcan, Sebnem, Luc Laeven, and David Moreno. 2018. "Debt Overhang, Rollover Risk, and Corporate Investment: Evidence from the European Crisis," NBER Working Papers 24555, National Bureau of Economic Research, Inc. - Organization for Economic Cooperation and Development (OECD). 2017. “Business Dynamics and Productivity” OECD Publishing, Paris. - OECD. 2018. “Financing SMEs and Entrepreneurs 2018: An OECD Scoreboard.” OECD Publishing, Paris. *Italicized source: References list from apd-reo-background-papers-1018 (apd-reo-background-papers-1018 - References).* ### introduction of the Internet in the 1990s , and most recently, ### apd-reo-background-papers-1018 - introduction of the Internet in the 1990s , and most recently, ### Definition and measurement of the digital economy - Narrow definition: ICT sector or “digital sector,” including telecommunications, the Internet, ICT services, hardware and software. - Broad definition: ICT sector plus parts of traditional sectors integrated with digital technology; G20 definition: “a broad range of economic activities that includes using digitized information and knowledge as the key factor of production, and modern information networks as the important activity space.” - Measurement divergence: - Narrow OECD-based estimates: China’s digital economy at 6 percent of GDP, Japan’s at 8 percent, South Korea’s at 10 percent (Figure 1). - Broad China Academy of Information and Communication estimate (national accounts basis): China 30 percent of GDP, Japan 46 percent, United States 59 percent (Figure 2). - Paper focus: primarily the narrow definition due to data availability, with additional measures for robotics and e-commerce and a blended index created for the paper. ### Supply: Digital dividend and ICT sector in GDP - Asia’s ICT presence: - In 2015, seven of the world’s top 10 economies in terms of ICT share of GDP were from Asia (Figure 3). - Taiwan Province of China ICT/GDP ratio rose from 9.3 to about 18 percent of GDP during 2005–15. - Korea’s ICT/GDP ratio increased from 7.5 to about 10 percent of GDP during 2005–15. - ICT growth relative to GDP (2005–15): - ICT sector grew twice as fast as GDP in India and Thailand. - ICT grew about 40 percent faster than GDP in China. - ICT grew nearly four times as fast as GDP in Japan (Figure 4). - Components of ICT growth vary by economy: - Taiwan Province of China, Korea, China: ICT manufacturing (electronic components) dominant. - India: computer-related services account for about 70 percent of ICT value added; telecommunications also important. ### Trade in digitalization and ICT exports - Asia accounts for about half of global ICT exports; within Asia, China’s contribution is about 60 percent (including Hong Kong SAR) and has driven more than half of ICT export growth over the past decade, followed by South Korea (Figure 5). - ICT exports composition: majority goods exports; services gaining momentum. - Weak correlation between income level and ICT exports: - India: ICT services constitute almost 70 percent of total exports despite per capita income below $2,000. - China: largest ICT exporter in gross terms but domestic value added in ICT about the same as in India or the United States, reflecting assembly-hub role in global value chains (Figure 7). ### Use of digital technologies: digital divide and convergence - Asia shows the highest dispersion in digital adoption: - Trendsetters: Korea, Japan. - Low adopters: Myanmar, Lao P.D.R. - Rapid adopters: Bangladesh, Cambodia. - At any given income level, Asian economies generally adopt digitalization more than global peers (Figure 8). - Digital convergence: - Lower-adoption economies are catching up; 2012–16 comparison shows Asian advanced economies with relatively low growth in digital usage (already high starting levels), while several ASEAN and low-income Asian countries gained significant momentum (Figure 9). - Digital User’s Index composition: average of six indicators (mobile phone subscriptions per 100 population; percentage of individuals using the Internet; percentage of households with a personal computer; percentage of households with Internet access; fixed broadband subscriptions per 100 population; mobile-broadband subscriptions per 100 population). ### Use of robotics - Over 60 percent of the world’s industrial robots are used in Asia; China dominant with twice as many robots as Korea (Figure 10). - Robot density in manufacturing, 2016 (number of industrial robot stock per 1,000 employees): - World average: about 7.4 robots per 1,000 employees in manufacturing (global average in 2016). - Korea: 63 - Singapore: 49 - Japan: 30 - Emerging markets such as Thailand and India are among top users of industrial robots; robot adoption in Asia continues to grow faster than in other regions. ### Fintech - Digital payments: Asia second only to Europe in digital payments (Figure 12). - Thailand: - Government-backed electronic money transfer service prompted banks to waive fees for retail e-transactions in April 2018. - Mobile banking is replacing Internet banking (Figures 13 and 14). - Commercial bank branches in Thailand shown with year-over-year percent change and units (Figure 13). - Transaction volumes: Internet banking vs Mobile banking (millions of units) show mobile banking growth overtaking Internet banking (Figure 14). - Indonesia and Malaysia: e-money gaining ground; Indonesia’s e-money transactions show large growth (Figure 15; measured in trillions Indonesian rupiahs). - Banking implications: - Fintech can significantly reduce costs and time in cross-border transactions, increasing banks’ efficiency. - Example: Singapore’s DBS reported digital customer businesses were 42 percent more profitable than traditional channels, with the highest margins on Internet-based services. - ASEAN developments: Malaysia’s CIMB group planned launches of full-digital banks in Vietnam and the Philippines in 2018; Singapore’s DBS started India’s first mobile-only bank in 2016 and launched Digibank in Indonesia in 2017. ### E-commerce - Asia leads regionally in e-commerce sales as a share of retail sales (Figure 16). - Selected e-commerce sales (percent of retail sales): - China leads (Figure 17); South Korea, Japan, India among top 10 globally. - Indonesia: e-commerce sales as percent of total sales increased fourfold between 2014 and 2017 (Figure 18). - E-shopper penetration (percent of online population that bought a product online) is lowest in Asia relative to other regions, indicating significant untapped potential (Figure 19). ### Public finance and government digital adoption - Government digital adoption: Asian economies span the gamut; overall lag behind North America and Europe (Figure 20). - Top three global leaders in digital adoption are from Asia: South Korea, Singapore, Japan. - Malaysia, India, China perform better than the European average on the Digital Adoption Index. ### Innovation and computing power - Innovation tilt: Asia specialized in ICT patents; top five economies in ICT share of patents are China, Korea, Taiwan Province of China, India, Malaysia (Figure 21). - Among top 10 technology companies by market capitalization, only three are in Asia; R&D expenditures of these companies are below competitors in other regions (Figure 22). - China overtook the United States in terabytes processed in a minute as early as 2014 and is catching up in supercomputers and number of artificial-intelligence enterprises (Figures 23 and 24). - AI enterprises, 2016: counts show USA 1,078; China 597; Britain 135; Israel 74; Canada 70; and other country breakdowns (Figure 23). - Distribution of most powerful supercomputers, 2017: China 206; United States 124; Japan 36; United Kingdom 22; Germany 21; etc. (Figure 24). ### Asia’s growth and the digital sector: accounting and results - Framework: growth in labor productivity (y*) decomposed into capital intensity (K/Y), rising labor quality (h), and growing TFP proxied by R&D intensity in ICT and non-ICT sectors and population growth, following Jones (2002) and extensions. - Historical context: earlier literature emphasized factor accumulation for Asia’s growth in 1960s–1980s; this paper finds a shift toward technological progress since the 1990s. - R&D and patents: - R&D-expenditure-to-GDP ratio and share of researchers in total employment (R&D intensity) higher in Japan, Korea, Taiwan Province of China than in the United States (Figures 25 and 26). - Asia’s share of triadic patent families increased to 43 percent over past 15 years, driven by China and Korea (Figure 27). - Share of researchers working on digital sector: China 18 percent; Australia 23 percent; Japan 34 percent; Korea 46 percent; Taiwan Province of China 73 percent; OECD average 30 percent (Figure 28). - Empirical results, 1995–2016: - TFP growth (technological progress) explains most economic growth over 1995–2016, with heterogeneity across countries. - Human capital contribution: increases in human capital contributed 11.9 percent to per capita income growth on average; range from -0.5 percent in New Zealand to 27 percent in Singapore. - Capital deepening: negative contribution in some economies affected by the Asian financial crisis, ranging from -78.6 percent in Thailand to -7.2 percent in Indonesia (Figure 29). - Digital sector innovation contribution: around 28 percent of per capita growth over 1995–2016, with contributions ranging between 12 percent (Singapore) and 49 percent (Thailand). Note: analysis uses narrow (OECD-based) definition; share could be underestimated under a broader definition. - ICT capital deepening: - Korea evidence: ICT capital deepening increased while overall capital deepening contribution was negative; Figure 30 shows Korea’s capital deepening changes for Information and communication technology (ICT), Non-ICT, and Overall (percent change; year-over-year average). - Japan: real fixed investment in computer software and ICT equipment grew at annual average rates of 4.2 percent and 5.2 percent, respectively, over 1995–2016, versus -0.1 percent for overall fixed investment. - Singapore: intellectual property product capital stock (9 percent of net capital stock as of 2017) grew twice as fast as overall capital stock during 1995–2016, increasing its share in real GDP to nearly 25 percent. - Projection: - Assuming current trends continue, innovation in the digital sector could account for 36 percent of Asia’s economic growth within 15 years. *IMF staff background paper (apd-reo-background-papers-1018) excerpt.* ### Conclusion ### Conclusion ### Overall contribution of the digital sector to Asia’s per capita growth - Technological progress is the main driver of Asia’s per capita growth over the past 25 years. - Digital technological progress accounts for between 12 and 49 percent of per capita growth. ### Automation and the Future of Work in Asia - Scope and context - In 2016, there were about 1.8 million industrial robots operating in the world. - More than half of the estimated operational stock of industrial robots is in Asia (1 million units out of a total of 1.8 million units in 2016). - Industrial robots are used almost exclusively in manufacturing, with automotive manufacturing being by far the most automated subsector. - The analysis focuses on industrial robots because of data availability (automation now spans AI, drones, self-driving cars, distributed ledger technologies, smart contracts, and decentralized autonomous organizations, but robots have richer data). - Key empirical findings (sample and period) - Sample covers 14 manufacturing subsectors in 40 economies in Asia, Europe, and the Western Hemisphere for the period 2010–14. - Robot penetration is not significantly associated with net employment losses in the full sample of 40 economies. - Restricting attention to Asia, increased robot usage is associated with lower employment growth in manufacturing: - One more robot per 1,000 employees is associated with a 0.26 percentage point decrease in employment growth in manufacturing sectors (Asia). - The negative employment effect in Asia is driven by highly automated sectors and economies (manufacturing of automotive components, plastic and rubber products, electronics) where robot density was relatively high in 2010 and increased rapidly thereafter. - Employment data used do not capture jobs created outside the sectors (for example, robotics repair and maintenance services). - Country-specific heterogeneity: Japan as an example - Japan (prefecture-level analysis): higher exposure to robots is associated with higher productivity and employment growth. - An increase of robot density by 1 percent corresponds to a 15 percent increase in TFP growth for all samples. - The same 1 percent increase corresponds to a 6 percent increase in TFP growth in a manufacturing subsample. - A 1 percent increase in robot density leads to a 0.2 percent increase in employment growth. - Japan’s experience suggests economies with declining labor forces (for example, China, Korea, Thailand in similar demographic scenarios) may benefit from automation. - Distributional impacts by skill/education - Automation disproportionately affects routine and manual tasks prevalent in manufacturing. - Penetration of industrial robots is negatively related to employment growth for workers with secondary education; no significant relation for those with higher education. - Quantified effect: a one standard deviation increase in robot penetration at the economy level (equivalent to about 0.12 more robots per 1,000 employees in an economy over 2010–14) is associated with a decrease in employment of about 0.24 standard deviations (or about a 0.01 percentage point decrease in employment in the sample) on average across sample economies. - Interpretation and policy implications - Automation brings both opportunities (productivity gains, mitigation of demographic headwinds) and challenges (displacement of workers, especially middle-skill). - A “critical mass” of robots may be needed before sectoral impacts become apparent; productivity gains may require complementary innovations and redesign of production processes and occupations. - Productivity effects may follow a “J-Curve,” with productivity possibly declining before ultimately increasing. - Policy recommendations: - Provide training and re-training opportunities to help vulnerable workers adapt and acquire skills in demand. - Implement policies that help create more flexible labor markets, including active labor market policies to absorb employment displacement related to automation. ### E-commerce as a New Engine for Growth - Firm-level and macro linkages - Participation in online commerce is associated with a more than 30 percent increase in TFP at the firm level in Asia. - Firms engaged in e-commerce export 50 percent more. - E-commerce appears especially beneficial for small firms in Asia. - E-commerce can boost private consumption and investment by improving access to a wider range of products and services at lower prices. - Evidence on new consumption from country studies - China (McKinsey study): out of $100 in Internet spending, close to 40 percent represents incremental (new) consumption while the remaining 60 percent is diverted from traditional offline retail channels. - Indonesia (McKinsey study): about 30 percent of online commerce spending is new consumption, capturing previously untapped needs. - Potential channels - For consumers: better access, lower prices → boost consumption. - For firms: new business opportunities, access to larger markets → support investment. - Cross-border e-commerce growth can increase participation in regional and global value chains and support international trade. ### Evidence from firm-level data (World Bank Enterprise Surveys) - Sample and definition - WBES sample covers 78 developing economies across all regions, including six Asian economies (China, Indonesia, Mongolia, Nepal, the Philippines, and Vietnam); surveys carried out between 2006 and 2012. - E-commerce is a dummy variable equal to one if a firm uses an Internet connection to order purchases from other establishments or to deliver products and services to clients. - Performance and characteristics of e-commerce firms - Firms engaged in e-commerce tend to have: - A more educated labor force. - Better access to finance. - Greater innovation activity (new products/processes, R&D, technology licensed from foreign companies, internationally recognized quality certifications). - Selected comparative statistics (E-commerce firms vs Other firms, all countries / Asia where given): - Firms owning a website: 70.9% vs 53.6% (All Countries). - Communicate by e-mail with clients and suppliers: 97.5% vs 82.9% (All Countries). - Use of technology licensed from a foreign company: 21.2% vs 13.4% (All Countries). - New products/services introduced during last 3 yrs: 59.1% vs 45.0% (All Countries). - Did the firm spend on R&D during last fiscal year: 50.4% vs 36.4% (All Countries). - Workers with completed high school education: 73.0% vs 63.7% (All Countries). - Firms with line of credit or loan: 53.9% vs 42.8% (All Countries). - Annual sales (median, in thousands of US dollars): 174311 vs 57269 (All Countries). - Value added (median, in thousands of US dollars): 56341 vs 8109 (All Countries). - Export in total sales (%): 6.7 vs 4.2 (All Countries). - In Asia, firms engaged in online activities have on average 50 percent higher labor productivity than other firms (potential outliers excluded: top 1 percent and bottom 1 percent). - TFP distribution and firm size - Firms with online activities have higher TFP, particularly in Asia. - E-commerce is especially beneficial for small firms in Asia. *IMF staff calculations based on World Bank Enterprise Surveys; International Federation of Robotics; World Input-Output Database; International Labor Organization.* ### 1. All Economies (All Firms) 2. Asia (All Firms) ### apd-reo-background-papers-1018 - 1. All Economies (All Firms) 2. Asia (All Firms) ### Methodology: Firm-level Productivity Estimation - Total Factor Productivity (TFP) for each firm i in country c, sector s, time t (V_csit) is estimated as a residual from an equation explaining value added by: - number of total employees (L_csit) and - level of capital stock (K_csit). - All variables are in logarithm terms. - All regressions include country and sectoral dummies. - Firm-level estimations control for firms' age, size, foreign ownership, and export status. ### Labor and Firm Productivity Findings - Labor productivity is measured as the ratio of value added to number of employees. - Controlling for firm characteristics confirms prior results: firms involved in e-commerce are more productive. - The potential impact of e-commerce on firm productivity seems to be greater in Asia than in other developing regions. - Estimated impacts (Figure 37, note): - For Asia, participation in e-commerce is associated with more than a 30 percent increase in total factor productivity. - For Asia, participation in e-commerce is associated with an increase in the share of exports to total sales by about 2 units, corresponding to a 50 percent rise. - Significance indicators shown: ** p<0.05; *** p<0.01. ### Channels: Innovation, Human Capital, Access to Finance, and Exports - Drivers of higher productivity among online firms: - More highly skilled labor force. - Faster pace of innovation. - To some extent, better access to finance enabling delivery of products and services with internationally recognized quality certification. - E-commerce firms also generate a larger share of sales revenues from export, particularly in Asia. - A better skilled labor force and higher product quality appear to support higher exports by firms with online activities. - The role of skill premia in supporting export activities seems particularly important in Asia. ### Platforms: Amplification of Benefits and Competition Concerns - Platforms can magnify e-commerce benefits through network effects and positive externalities: - Increased competition within markets and pressure to lower prices (reduced search costs). - Broader geographical reach of suppliers and savings in supply chain management. - Access to advanced ICT infrastructure, data centers, applications, and processes for small and medium enterprises. - Same-side network effects: more providers or customers enhance platform efficiency and attract yet more users, aided by big data customization. - Examples cited: - Ant Financial Services Group (affiliate of Alibaba) collects information from Taobao to extend the credit frontier to firms not served by traditional banks. - Flipkart (India) implemented training (account management, marketing, warehousing), dedicated supply chain and last-mile infrastructure, and a light e-commerce application functioning with limited Internet bandwidth or offline. - Competition and regulatory concerns: - Economies of scale and exclusive access to information by large platforms pose anti-competitive concerns. - Network effects make switching platforms challenging, reinforcing platform market power and risk of anti-competitive practices. - Policy areas requiring attention include taxation, competition, and data privacy. - The European Union view cited: platforms are too diverse for a one-size-fits-all solution and merits case-by-case application of antitrust law (Lougher and Kalmanowicz 2016). - Concluding note: Further research and regulatory experiments are needed to establish a formal framework for the platform economy. ### Digitalization of Finance (Fintech) in Asia - Fintech can support growth and poverty reduction by strengthening financial development, inclusion, and efficiency, but also poses risks to the financial sector. - Regional adoption and funding: - In Asia, digitalization of finance has been growing faster than the global average. - Three of the five economies identified as having the highest rate of Fintech adoption globally are in the region: China, India and Australia. - Since 2010, investments have picked up, led by China but also in Southeast Asia by Singapore, Malaysia and Thailand (IMF 2018b). - Cumulative Fintech equity funding reached about $28 billion in 2017, with two-thirds of that growth captured by China. - Heterogeneity in fintech development: - Economies have adopted a wide range of technologies based on consumer needs, level of development, regulatory stance, and existing financial and technological infrastructure. - Examples: rapid growth of mobile payments in China versus growth of contactless card payments in Australia. - Some economies did not develop mobile money products monetizing pay-as-you-go phone credit because “postpaid” monthly phone contracts became standard. - Empirical evidence: - Analysis using data on digital payments between 2014 and 2017 finds economies catching up to the frontier of universal access to digital payments. - Economies with low levels of digital payment use in 2014 have significantly higher growth rates (convergence evidence). *Source: apd-reo-background-papers-1018 - 1. All Economies (All Firms) 2. Asia (All Firms), IMF staff calculations and World Bank Enterprise Surveys.* ### Chapter 1 of the April 2018 GFSR. ### apd-reo-background-papers-1018 - Chapter 1 of the April 2018 GFSR ### Fintech, digital payments, and financial inclusion - Econometric evidence indicates that digital financial services can boost financial inclusion. - Asia: nearly 30 percent of the population still lacks access to even a basic savings account. - Complementary benefits: Fintech facilitates other digital activities such as e-commerce and can translate into positive macroeconomic outcomes through greater inclusion. - Convergence in digital payments use observed over 2014–2017 despite heterogeneity in technologies and business models. - Regional patterns: - Europe and the Western Hemisphere: pronounced negative association between leapfrogging to cellular technologies and traditional financial infrastructure (notably bank branches). - Africa and the Pacific: mixed evidence where technology may complement traditional financial service delivery even after controlling for income and cellular access. ### Who uses digital financial services — distributional patterns and gaps - Usage has replicated conventional finance patterns, with disparities by gender and income: - Bangladesh: 20 percent report having a mobile money account; men 30 percent, women 10 percent. - Regional disparities illustrated by figures showing access by richest 60% vs poorest 40%, and by male vs female populations (Global Findex data). - Risk of a digital divide without targeted policy attention; unequal access and technological literacy can exacerbate inequality. ### Financial-sector risks from Fintech - Key risks identified: - Online payment service providers operating in the shadow banking system: pooling cash and investing in interbank certificates of deposits or micro loans, creating credit and maturity transformation outside regulatory supervision. - Capital-light consumer lending models among Fintech firms: potential for large consumer losses and liquidity-driven credit risks. - Weak “know your customer” standards: risks of embezzlement, fraud, and AML/CFT concerns due to limited clarity on transaction size/nature and opaque user identities. - Close integration across financial-service segments: potential for amplified spillovers and losses in downturns. - Structural considerations: - Rise of large hybrid technology/financial companies could produce sizable spillover effects. - Fragmentation of services along the payments chain may reduce controls over data processing and risk management. ### Digitalization to strengthen public finance — taxation, methodology, and quantitative results - Methodology (summary): - Uses estimates from Chapter 2 of the April 2018 Fiscal Monitor. - Quantifies average gains from reducing the gap with the digitalization frontier by 50 percent, measured by the UN Online Service Index (scope and quality of public sector online services). - Estimates impact on misreporting of import prices via a gravity model with country and time fixed effects and controls, then computes increases in VAT and tariff revenue using implied higher reported import prices and applicable tax rates. - Also estimates increases in wealth and income taxes related to undeclared offshore wealth using tax rates and country-specific offshore wealth estimates. - Potential import-VAT revenue gains (closing half the distance to digitalization frontier): - Asia: 0.6 percent of GDP. - ASEAN: 1.2 percent of GDP. - Pacific island countries: 2.5 percent of GDP. - Median gains: advanced 0.1 percent of GDP; emerging 0.7 percent of GDP; low-income 1.8 percent of GDP. - Interpretation caveat: digitalization correlates with GDP per capita and Asian countries are closer to the frontier for each income bracket, so smaller estimated gains may reflect smaller gaps to the frontier. - Potential tariff revenue gains: - Asia average: 0.2 percent of GDP. - Pacific island countries in Asia: 0.7 percent of GDP. - ASEAN: 0.5 percent of GDP. - Median gains: emerging markets 0.2 percent of GDP; low-income countries 1.1 percent of GDP. - Potential revenue gains from offshore wealth: - Asia: 0.2 percent of GDP. - Estimated offshore wealth of Asian economies: 7.3 percent of GDP. - South Asia: estimated tax increase 0.3 percent of GDP. - Within Asia, advanced economies have slightly higher median estimated gains despite lower proportions of offshore wealth compared with some other income groups. ### Risks to revenue and base erosion from digital platforms; examples of policy responses - Risks: - Base erosion via shifting transactions to lower-tax sectors, informal sector, or abroad (P2P platforms and e-commerce can facilitate this). - Internationalization of transactions can reduce domestic tax bases. - Examples of policy responses and measures: - Withholding and reporting by platforms: India requires digital platforms to charge and remit service taxes due on sellers' income; Australia requires drive-sharing platform drivers to register as businesses and charge goods and services tax. - Multilateral efforts: OECD BEPS Project includes the digital economy; implementation ongoing. - Specific national measures: India’s “equalization levy” requires a 6 percent withholding on specified nonresident digital services related to online advertising; Australia and the United Kingdom introduced “diverted profits tax” targeting profits diverted to low-tax jurisdictions. ### Digitalization and better public spending: social safety nets and digital ID - Asia's public spending on social safety nets: 1.2 percent of GDP (lower than developing Europe, Latin America and Caribbean, and Sub-Saharan Africa). - Digital social registries: - Support outreach, intake, registration, and eligibility determination for social programs. - Philippines’ Listahanan registry: registration gateway for as many as 52 social programs, with 75 percent of the population registered. - Social registries often rely on self-reported data; reducing inclusion errors requires verification with civil and land registries. - Digital identification: - Digital ID systems store personal data and digital credentials for authentication and can reduce duplicates and “ghost” beneficiaries. - Many developing Asian economies already have operationalized digital ID systems. - India’s Aadhaar: 12-digit biometric ID for 1.2 billion residents; linked to social programs and used to reduce subsidy leakage in the liquefied petroleum gas program through beneficiary authentication and direct electronic transfers to Aadhaar-linked bank accounts. - Cautions: - Digital ID and registries require data verification, privacy protections, cybersecurity, and fallback options to avoid exclusion of genuine beneficiaries. ### Policies to harness digital dividends and manage risks - Policies to facilitate technological advances: - Enhance productivity, encourage R&D in digital and other sectors, promote diffusion of global knowledge, upgrade physical and soft infrastructure, and improve access to and quality of education. - Increase R&D intensity via patent policy, competition in research grants, and optimal government subsidies. - Invest in human capital to build innovation absorption capacity. - Fostering e-commerce — enabling factors to address: - Economic prerequisites: Internet access, secure servers, payment methods (credit card, e-wallet, mobile payment), and reliable delivery services. - Legal and institutional environment: e-transaction laws, consumer protection, privacy/data protection, and cybercrime legislation; enforcement and competition policy capacity. - Noted gaps: Pacific island countries practically lack legislation covering electronic transactions, consumer protection, data protection, and cybercrime. - Policies to manage labor-market transitions and reduce inequality: - Revamp education to emphasize foundation skills, digital literacy, high-order thinking, and social/emotional skills. - Provide training and re-training, lifelong learning, and initiatives like Singapore’s SkillsFuture. - Use income redistribution and safety nets to soften impacts; Korinek and Stiglitz (2018) suggest supportive transition policies improve welfare outcomes while accommodating faster innovation. - Digitalization of finance — regulatory approaches: - Need international collaboration to share best practices. - Better data for monitoring Fintech developments and greater regulatory agility are required. - Promote competition and maintain level playing field while encouraging low leverage among new entrants. - Address AML/CFT and cybersecurity threats proactively. - Example of closing regulatory loopholes: China’s money market fund Yuebao now captured in People’s Bank of China money supply statistics. - Policies to strengthen public finance through digitalization: - Leverage platform data and withholding mechanisms to formalize transactions and collect indirect taxes. - Improve international information sharing: automatic exchange of nonresident financial account information (OECD/G20) and international registers of asset ownership and shareholders. - Consider new tax designs (for example, dual income tax systems linking capital income and wealth) facilitated by better data. - Safeguards and design considerations: - Mitigate cybersecurity risks and protect privacy in large government-run databases; legal and technological frameworks should limit access to minimal necessary information and ensure robust security. - Ensure digital inclusion or provide fallback options to prevent exclusion from social programs, including institutional and human capacity for Internet connectivity. - Balance enabling innovation with addressing digitalization-linked risks through comprehensive policy design. *Source: Chapter 1 of the April 2018 Global Financial Stability Report (apd-reo-background-papers-1018).* ### Box 1. Macroeconomic Implications of Automation: The Case of Japan ### Box 1. Macroeconomic Implications of Automation: The Case of Japan ### Methodology and Measurement - Prefecture-level robot density is estimated as a weighted sum of the share of robots used in each sector out of the total number of robots used in the economy, weighted by the share of a prefecture’s output contribution to a specific industry to the total output of the industry. - Notation and inputs used: - 푦푦푠푡푝: total output of industry s of prefecture p at time t. - 푦푦푠푡: total output of industry s at time t. - 퐻퐻푟푟퐻퐻 푟푟푖푖 푤푤푟푟푟푟푠푠푖푖푑푑푠푠푡푡: total hours worked. - 푟푟푟푟푟푟푟푟 푖푖 푖푖푖푖푟푟 푠푠푠푠푡푡: operational stock of robots for industry s at time t. - Robot density varies across prefectures depending on industrial composition; prefectures with the highest robot density have large automotive and electronics manufacturing plants (examples: Aichi, Kanagawa). Prefectures with the lowest robot density include Akita and Aomori. ### Main Empirical Findings - Panel regressions with prefecture and industry group fixed effects, controlling for the global financial crisis, indicate: - For the manufacturing sample: - An increase of robot density by 1 percent corresponds to 6 percentage increase in total factor productivity (TFP). - A 1 percent increase in robot density leads to 0.3 percentage point increase in employment growth. - Spillover to services: - Higher robot density per prefecture is significantly correlated with higher TFP growth of the service sector: a 1 percent increase in robot density is associated with higher service TFP. - However, higher automation in manufacturing is negatively correlated with some labor market outcomes in service sectors of that prefecture: - Controlling for crisis and prefecture fixed effects, an increase of robot density by 1 percent leads to a 1.2 percentage point decrease in employee growth in the service sector. - The impact on service sector wage growth is negative, but insignificant. - Table 1.1 (Impact of Robots on Manufacturing and Service Sector Productivity and the Labor Market at the Prefecture Level) — selected coefficient estimates and standard errors (standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1): - Manufacturing: - Robot density (prefecture): 6.308*** (2.657) for TFP - Robot density (prefecture): 0.0912 (0.171) for Wage Growth - Robot density (prefecture): 0.275*** (0.078) for Employee Growth - Crisis*Robot density: -14.28*** (0.996) for TFP - Crisis*Robot density: -0.533*** (0.064) for Wage Growth - Crisis*Robot density: -0.173*** (0.029) for Employee Growth - Constant for TFP: -22.63 (22.63) - Observations: 940; R-squared (TFP): 0.207 - Services: - Robot density (prefecture): 3.235*** (0.911) for TFP - Robot density (prefecture): -2.132 (1.905) for Wage Growth - Robot density (prefecture): -1.208*** (0.433) for Employee Growth - Crisis*Robot density: -1.159*** (0.353) for TFP - Crisis*Robot density: -3.687*** (0.737) for Wage Growth - Crisis*Robot density: -1.136*** (0.167) for Employee Growth - Constant for TFP: -1.789** (0.826) - Observations: 940; R-squared (TFP): 0.048 - Note: Prefecture fixed effects included for all regressions. ### Distributional Concerns and Labor Market Implications - Automation may disproportionately affect female and low-skilled workers: - Existing studies indicate industrial robots have contributed to increased inequality by replacing middle- or low-skilled jobs that are more automatable. - Hamaguchi and Kondo (2017) estimate “computerizability” of occupations and find female workers are exposed to higher risks of computerization because they tend to be engaged in occupations with higher probability of computerization (examples: receptionists, clerical workers, sales workers). ### Policy Implications and Recommendations - Encourage automation in areas most affected by labor shortages due to demographic challenges, highlighting the positive sides of automation (Schneider, Hong, and Lee 2018). - Advance technology use of automation in non-manufacturing sectors, especially age-related services that face increasing automation needs. - Provide adequate training and education to improve skill quality and adaptability: - Refocus general education on problem-solving and emphasize cognitive skills to help workers shift across occupations (IMF 2018c). - Implement policies to encourage participation of untapped labor force segments (such as female workers) and consider the role of foreign workers to offset demographic challenges. *Source: IMF staff calculations; International Federation of Robotics; Japan Statistical Federation. Box prepared by Gee Hee Hong and Todd Schneider of the IMF’s Asia and Pacific Department.* --- _Source: https://www.imf.org/-/media/files/publications/reo/apd/2018/apd-reo-background-papers-1018.pdf_