## wp1916

## Source details

**Canonical URL:** [wp1916](https://www.imf.org/-/media/files/publications/wp/2019/wp1916.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wp1916.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wp1916.pdf.json)

---

### Scale advantage and adopters
- 770 million 4G users in China, accounting for 58 percent of total mobile users.
- 700 million internet users and 282 million digital natives (internet users that are less than 25 years old).
- India had internet users about 60 percent of China’s size in 2016.
- United States has less than 300 million users.
- Ecosystem developed by “BAT” (Baidu, Alibaba and Tencent) leverages multi-industry reach and rapid accumulation of consumer data.

### Evolution of the digital economy
- Digital economy size increased from 15 percent of GDP in 2008 to 33 percent in 2017.
- New ICT sectors: overall size remains small and largely stable at around 7 percent of GDP.
- Digitalized traditional sectors expanded from 10 percent of GDP in 2008 to 25 percent in 2017.
- Based on the Fletcher digital adoption index, China’s speed of digitalization is the fastest in their 62-country sample.
- Rising penetration of broadband: fixed broadband users and fiber broadband user penetration trends documented (China Academy of Information and Communication Technology).

### Digitalization across provinces and sectors
- Provincial variation broadly in line with income level:
  - Beijing 43; Shanghai 44; Zhejiang 35; Guangdong 31; Tianjin 31; Jiangsu 30; Fujian 30; Shandong 29; Hubei 27; Chongqin 27; Sichuan 27; Liaoning 24; Hebei 24; Jiangxi 22; Henan 18 (share of digital economy as percent of GDP; GDP per capita in RMB).
  - Beijing and Shanghai: digital economy close to 45 percent of GDP.
  - Central Henan province: digital economy about 15 percent of provincial GDP.
- Sectoral digitalization (2017):
  - Service sector: ICT contributing to 33 percent of the sector’s value-added.
  - Industrial sector: ICT contributing 17 percent of its value-added.
  - Agriculture sector: ICT contributing 7 percent of digitalization.
- Within services, most ICT-intensive subsectors are in financial services and entertainment; within industry, advanced manufacturing is more digitalized.

### Productivity impacts
- Data availability and processing:
  - Amount of data available has been roughly doubling every two years.
  - Data processing capacity doubles every eighteen months.
- Empirical results:
  - Growth of the digital economy (broad CAICT definition) correlated with TFP growth with a coefficient of 0.5.
  - Regression: a 1 percentage point increase in the overall digitalization of the economy is associated with an increase of 0.3 percentage point of GDP growth (model 1 in appendix), with a two-year lag.
  - Province-level regressions (Tencent data) show a similar impact, somewhat higher in developed regions.
- Channels for efficiency gains:
  - Lower transaction costs (fintech/mobile transactions).
  - Reduced information asymmetry and better matching (platforms and big data analytics).
  - Enhanced production efficiency (automation: shorter cycle times, improved quality and reliability).

### Employment impacts
- Job creation:
  - Alibaba’s platform: almost 11 million SMEs, which have created over 30 million jobs over the past decade.
  - Didi taxi platform connected to 13 million drivers.
  - ICT sector: 1.4 million jobs added for high-skill workers in the past five years; average wage has doubled since 2012.
- Job losses and disruption:
  - Industrial employment fallen by 9 million since 2012.
  - Foxcom replaced 60 thousand workers with 40 thousand robots.
  - Retail disruption visible, though retail employment growth broadly stable despite surging e-commerce.
- Net employment impact (regression evidence):
  - Given GDP growth, a one percentage point increase in the digital economy has boosted employment growth by 0.01 percent, with a one-year lag.
  - Considering average growth of the digital economy was 10 percent, it has contributed 0.1 ppt to employment growth.
  - Aggregate employment growth was 0.2-0.3 ppt in the past few years, so digitalization has accounted for one third to half of total employment growth.
- Expected trends:
  - Overall employment growth rate expected to slow to below 0.1 percent.
  - Digitalization may increase labor market polarization, hollowing out mid-skilled jobs while raising wages and employment for high- and low-skill labor.
  - Downside risk if AI is applied on a large scale: potential for larger-than-anticipated labor market disruption.

### Market structure effects
- Three channels shaping market structure:
  1. Disintermediation: linking supply and demand directly via digital platforms.
  2. Increase of small players in traditional sectors: lower entry barriers and easy access to large consumer bases.
  3. Oligopoly in platform industries: higher barriers in digital platforms lead to dominance by a few firms (example: Alibaba and Tencent in mobile payment and broader financial services), creating economies of scale or “information scale” and potential for price distortions if competition is lacking.

### Digitalization and green development
- Digitalization promotes services via “super apps” (entertainment, dining, education, health).
- Digital tools enable carbon tracking:
  - Ant Financial’s Ant Forest mobile application: by end 2017, the app has 288 million users.
  - Cumulative avoided carbon emission reached 2.05 million tons and cumulative trees planted amounting to 13.14 million.

### Digitalization and inequality
- Targeted poverty reduction: more than 1300 “Taobao villages” link remote suppliers to consumer markets.
- Financial inclusion: digitalization contributed to increasing financial inclusion (World Bank 2017).
- Offsetting distributional effects:
  - Disruption borne mostly by relatively low-skilled workers; high-skilled workers likely benefit more.
  - China’s inequality has been on a modest downward trend post GFC, which has stalled in recent years.
- Key statistics:
  - Internet penetration in rural areas: 19 percent of the rural population.
  - World Bank (2016) estimate: 77 percent of employment in China is susceptible to automation.
  - Robot intensity in manufacturing:
    - China: 5 percent
    - U.S.: 18 percent
    - Korea: 60 percent
  - World Average robot density: 74 (installed industrial robots per 10,000 employees in manufacturing, 2016).

### Fintech developments and scale
- Fintech areas: third-party payments, peer-to-peer lending, internet credit including microlending, internet-based banking and insurance, digital wealth management, and credit-ratings.
- Large fintech players have built integrated ecosystems (Alibaba, Ant Financial, Tencent, JD.com, Baidu, Ping An).
- Third-party mobile payments:
  - Payments through third-party processors reached over RMB 119 trillion (or roughly US$18 trillion) in 2016 (PBOC/World Bank).
  - Alipay and WeChat Pay dominate with 84 percent of China’s market share.
  - Most mobile payments are no more than $20 per transaction.
  - Mobile payments made up about 75 percent of total payment in 2016.
  - In the U.S., 20 percent of e-commerce payments come from mobile phones (Goldman Sachs 2017).
- Alipay / Ant Financial:
  - Yuerbao launched in 2013.
  - By end-2017: number of users 474 million; funds under management RMB 1.6 trillion, 2.5 percent of total bank deposits.
- Internet banks and micro-lending:
  - Three internet banks highlighted: WeBank, MyBank and XW Bank.
  - WeBank products: unsecured microloans (Weilidai) and auto microloans (Weichedai) that can lend up to RMB 200,000 without collateral or guarantee.

### P2P lending and internet-based credit
- P2P platforms act as information intermediaries; funding primarily from retail investors.
- Growth prior to regulatory tightening:
  - By end-2016 total P2P transaction volume reached RMB 2.06 trillion, more than double in just a year and equivalent of 12 percent of total bank loans extended in 2016 (PBOC).
  - Maturity of these loans: averaging 5 to 8 months.
  - Average rate of return: 10.5 percent in 2016, falling from over 13 percent a year ago (PBOC).
- Recent regulatory tightening has reduced the number of P2P platforms significantly.

### Financial stability implications
- Fintech ecosystem lowers costs and expands services but alters market structure and creates risks:
  - Online payment service providers operate in the shadow banking system, pooling cash from banks and investing in interbank CDs or micro loans, causing credit and maturity transformation that is difficult to monitor.
  - Capital-light consumer lending models: consumers could suffer significant losses if these companies face a liquidity crunch.
  - Weak KYC standards create potential for embezzlement and fraud; AML/CFT concerns due to opaque transaction sizes and user identities.
  - Close integration across financial services could exacerbate risk spillovers and amplify losses in a downturn.
  - Closed-loop platforms allow credit creation outside central bank purview; policy responses cited include capturing Yuerbao in PBC money supply statistics and establishing Wanglian centralized clearing system.

### Regulatory responses and initiatives
- Key measures and timeline:
  - Guiding Opinions from the PBC and nine other ministries in 2015 and Provisional Rules from CBRC and three other ministries in 2016.
  - Series of announcements from June 2015 to October 2016 defining regulatory responsibilities across “internet finance” activities.
  - New FinTech committee created in May 2017 to coordinate regulators and industry.
  - Domestic crypto-asset exchanges closed; PBC does not recognize ICOs or crypto tokens as payment instruments; trading of bitcoins in RMB banned.
  - Wanglian, a centralized clearing house for all third-party payments, established in March 2017.
  - PBC mandated centralized custody of all client funds from online payment service providers, requiring deposits of all client funds at the PBC.
- Remaining gaps:
  - Lack of clear framework on data ownership and data sharing between market participants and regulators.
  - Need for comprehensive fintech supervisory oversight and coordination across regulators.

### Prospects, scenarios, and macro implications
- Continued rapid digitalization anticipated given large internet user base, established ecosystem, and improving infrastructure.
- Projection:
  - Size of the digital economy in China is likely to reach close to 50 percent of GDP by 2025.
- Macro and labor-force implications:
  - Digitalization can moderate but not reverse the downward trend in China’s potential growth as the economy matures.
  - Shift from industry to services since 2012 expected to continue, putting downward pressure on potential growth because services productivity is lower than industrial productivity in China.
  - Industrial job losses likely; overall employment impact likely contained if workers transition to services.
  - Digitalization may increase labor market polarization; downside risk from large-scale AI deployment.

### Conclusions and policy recommendations
- Summary: Digitalization brings significant benefits (boosting productivity, promoting rebalancing, creating jobs in new sectors) and risks (disruption of traditional sectors, job losses in mid-skill manufacturing, financial stability concerns from fintech).
- Recommended actions:
  - Active labor market policy: retraining the labor force to smooth transitions.
  - Strengthen social safety nets for affected workers close to retirement.
  - Competition policy: promote competition, encourage data sharing, consider policies supporting open networks and foreign firm entry; note that local data storage and cross-border data transfer requirements may hinder global integration.
  - Comprehensive fintech supervisory oversight: coordinate regulators and industry, focus on “form over substance” regulation, use regulatory sandboxes, reflect fintech risks in liquidity buffers and capital adequacy, strengthen data gathering and KYC for third-party payments.
  - Data privacy and consumer protection: supervisors coordinate with relevant authorities; establish laws to protect consumer data privacy and jurisdictions over data sharing.
  - E-government: improve digitalization of public services; UN e-government index rank for China in 2016: 63rd out of 193 countries.
  - Continue support for digital infrastructure and internet penetration, especially in rural areas.
  - Education policy: increase focus on ICT education in universities and vocational schools.
  - Play a more active role in setting global standards on regulation, cyber security, global digital standards and governance.

*Source: wp1916 - 2.05 million tons and the cumulative trees planted amounting to 13.14 million.*

### 2016. There were 770 million 4G users in China, accounting for 58 percent of total mobile users, a higher

### wp1916 - 2016. There were 770 million 4G users in China, accounting for 58 percent of total mobile users, a higher

### Scale advantage and adopters
- 770 million 4G users in China, accounting for 58 percent of total mobile users.
- Large base: 700 million internet users and 282 million digital natives (internet users that are less than 25 years old).
- India (roughly the same population as China) had internet users about 60 percent of China’s size in 2016.
- United States has less than 300 million users.
- The ecosystem developed by “BAT” (Baidu, Alibaba and Tencent) leverages multi-industry reach and rapid accumulation of consumer data to provide easy access to new products and services to millions of users.

### Evolution of the digital economy in China
- Digital economy size increased from 15 percent of GDP in 2008 to 33 percent in 2017, mainly driven by integration of ICT with traditional sectors.
- New ICT sectors: overall size remains small and largely stable at around 7 percent of GDP.
- Digitalized traditional sectors expanded from 10 percent of GDP in 2008 to 25 percent in 2017.
- Based on the Fletcher digital adoption index, China’s speed of digitalization is the fastest in their 62-country sample.
- Rising penetration of broadband: fixed broadband users and fiber broadband user penetration trends documented (sources: China Academy of Information and Communication Technology).

### Digitalization across provinces and sectors
- Provincial variation broadly in line with income level.
  - Beijing and Shanghai: digital economy close to 45 percent of GDP (similar to Japan’s level).
  - Central Henan province: digital economy about 15 percent of provincial GDP.
- Provincial digitalization chart values (share of digital economy as percent of GDP; GDP per capita in RMB) include examples:
  - Beijing 43; Shanghai 44; Zhejiang 35; Guangdong 31; Tianjin 31; Jiangsu 30; Fujian 30; Shandong 29; Hubei 27; Chongqin 27; Sichuan 27; Liaoning 24; Hebei 24; Jiangxi 22; Henan 18.
- Sectoral digitalization (2017):
  - Service sector: ICT contributing to 33 percent of the sector’s value-added.
  - Industrial sector: ICT contributing 17 percent of its value-added.
  - Agriculture sector: ICT contributing 7 percent of digitalization.
- Within services, most ICT-intensive subsectors are in financial services and entertainment; within industry, advanced manufacturing is more digitalized.

### The real impact of digitalization — Productivity
- Data has become a new factor of production; amount of data available has been roughly doubling every two years; data processing capacity doubles every eighteen months.
- Empirical results:
  - Growth of the digital economy (broad CAICT definition) has been highly correlated with TFP growth, with a coefficient of 0.5.
  - Regression result: a 1 percentage point increase in the overall digitalization of the economy is associated with an increase of 0.3 percentage point of GDP growth (model 1 in appendix), though with a two-year lag.
  - Province-level regressions (Tencent data) show a similar impact, somewhat higher in developed regions.
- Channels through which digitalization improved efficiency:
  - Lower transaction costs (fintech/mobile transactions).
  - Reduced information asymmetry and better matched demand and supply (platforms and big data analytics).
  - Enhanced production efficiency (automation: shorter cycle times, improved quality and reliability).

### The real impact of digitalization — Employment
- Job creation in new sectors:
  - Alibaba’s platform: almost 11 million SMEs, which have created over 30 million jobs over the past decade.
  - Didi taxi platform connected to 13 million drivers.
  - ICT sector: 1.4 million jobs added for high-skill workers in the past five years; average wage has doubled since 2012.
- Job losses amid digital disruption:
  - Industrial employment fallen by 9 million since 2012 (overcapacity cuts and automation-driven upgrading).
  - Foxcom replaced 60 thousand workers with 40 thousand robots.
  - Service-sector disruption most visible in retail, though retail employment growth has been broadly stable despite surging e-commerce penetration.
- Net employment impact:
  - Regression: given GDP growth, a one percentage point increase in the digital economy has boosted employment growth by 0.01 percent, with a one-year lag.
  - Considering average growth of the digital economy was 10 percent, it has contributed 0.1 ppt to employment growth.
  - Aggregate employment growth was 0.2-0.3 ppt in the past few years, so digitalization has accounted for one third to half of total employment growth.

### The real impact of digitalization — Market structure
- Three channels shaping market structure:
  1. Disintermediation: reduced layers of distribution, linking supply and demand directly through digital platforms.
  2. Increase of small players in traditional sectors: lower entry barriers allow small businesses easy access to large consumer bases at low cost.
  3. Oligopoly in platform industries: higher barriers in digital platforms lead to dominance by a few firms (example: Alibaba and Tencent in mobile payment and broader financial services), creating economies of scale or “information scale” but potential for price distortions if competition is lacking.

### Economic rebalancing and green development
- Digitalization promotes development of the service industry via “super apps” offering a one-stop shop for entertainment, dining, education, health, and other services.
- Digitalization can promote green development by making it easier to track carbon emissions.
  - Example: Ant Financial’s Ant Forest mobile application integrates carbon emission tracking with users’ daily consumption activities; by end 2017, the app has 288 million users, with the cumulative avoided carbon emission reaching

*Source: https://www.imf.org/-/media/files/publications/wp/2019/wp1916.pdf*

### 2.05 million tons and the cumulative trees planted amounting to 13.14 million.

### wp1916 - 2.05 million tons and the cumulative trees planted amounting to 13.14 million.

### Digitalization and inequality
- Digitalization can help targeted poverty reduction by linking suppliers in remote regions to consumer markets; example: more than 1300 “Taobao villages” in China.
- Digitalization contributed to increasing financial inclusion by providing easy mobile access of various financial services to rural residents (World Bank 2017).
- Offsetting effect: disruptive impact borne mostly by relatively low-skilled workers while high-skilled workers likely benefit more, potentially widening inequality.
- China’s inequality has been on a modest downward trend (which has stalled in recent years) post GFC despite rapid digitalization.
- Key statistics:
  - Internet penetration in rural areas: 19 percent of the rural population.
  - World Bank (2016) estimate: 77 percent of employment in China is susceptible to automation.
  - Robot intensity in manufacturing:
    - China: 5 percent
    - U.S.: 18 percent
    - Korea: 60 percent
  - World Average robot density: 74 (installed industrial robots per 10,000 employees in manufacturing, 2016).

### Fintech developments and scale
- Fintech areas in China include: third-party payments by non-bank digital providers, peer-to-peer lending, internet credit including microlending, internet-based banking and insurance, digital wealth management, and credit-ratings.
- Large fintech players have built integrated ecosystems linking customers with businesses across the finance supply chain (examples cited: Alibaba, Ant Financial, Tencent, JD.com, Baidu, Ping An).
- Third-party mobile payments:
  - Payments made through third-party processors reached over RMB 119 trillion (or roughly US$18 trillion) in 2016 (PBOC/World Bank).
  - Alipay and WeChat Pay now dominate with 84 percent of China’s market share.
  - Most mobile payments are no more than $20 per transaction.
  - Mobile payments made up about 75 percent of total payment in 2016.
  - In the U.S., 20 percent of e-commerce payments come from mobile phones (Goldman Sachs 2017).
- Alipay / Ant Financial:
  - Yuerbao launched in 2013.
  - By end-2017: number of users 474 million; funds under management RMB 1.6 trillion, 2.5 percent of total bank deposits.
- Internet banks and micro-lending:
  - Three internet banks highlighted: WeBank, MyBank and XW Bank.
  - Example product: WeBank offers unsecured microloans (Weilidai) and auto microloans (Weichedai) that can lend up to RMB 200,000 without collateral or guarantee.

### P2P lending and internet-based credit
- P2P platforms act as information intermediaries matching borrowers and lenders; funding primarily from retail investors.
- Growth prior to regulatory tightening:
  - By end-2016 total P2P transaction volume reached RMB 2.06 trillion, more than double in just a year and the equivalent of 12 percent of total bank loans extended in 2016 (PBOC).
  - Maturity of these loans: averaging 5 to 8 months.
  - Average rate of return: 10.5 percent in 2016, falling from over 13 percent a year ago (PBOC).
- Recent regulatory tightening has reduced the number of P2P platforms significantly.

### Implications for financial stability
- Large fintech firms have created universal banking and a fintech ecosystem, lowering costs and expanding services but also altering market structure.
- Key financial stability concerns and channels for spillovers:
  - Online payment service providers operate in the shadow banking system, pooling cash from banks and investing in interbank CDs or micro loans, resulting in credit and maturity transformation that is difficult to monitor.
  - Most fintech companies engaged in consumer lending follow a capital-light model; consumers could suffer significant losses if these companies face a liquidity crunch.
  - Weak KYC standards creating potential for embezzlement and fraud; AML/CFT concerns due to opaque transaction sizes and user identities.
  - Close integration across financial services could exacerbate risk spillovers and amplify losses in a downturn.
  - Closed-loop platforms allow credit creation outside central bank purview; policy responses cited include capturing Yuerbao in PBC money supply statistics and establishing Wanglian centralized clearing system.

### Regulatory responses and initiatives
- Regulatory timeline and measures:
  - Guiding Opinions from the PBC and nine other ministries in 2015 and Provisional Rules from CBRC and three other ministries in 2016 set regulatory framework for fintech credit.
  - Series of announcements from June 2015 to October 2016 defined regulatory responsibilities across “internet finance” activities, emphasizing compliance, funding models, and consumer protection.
  - New FinTech committee created in May 2017 to coordinate between financial regulators and industry participants.
  - Domestic crypto-asset exchanges closed; PBC does not recognize ICOs or crypto tokens as payment instruments; trading of bitcoins in RMB banned.
  - Wanglian, a centralized clearing house for all third-party payments, established in March 2017 to clear all third-party payments.
  - PBC mandated centralized custody of all client funds from online payment service providers, requiring deposits of all client funds at the PBC.
- Remaining regulatory gaps:
  - Lack of clear framework on data ownership and data sharing between market participants and regulators.
  - Need for comprehensive fintech supervisory oversight to close loopholes and coordinate across regulators.

### Prospects, scenarios, and macro implications
- Continued rapid digitalization anticipated given large internet user base, established online ecosystem, and improving digital infrastructure.
- Projection:
  - Size of the digital economy in China is likely to reach close to 50 percent of GDP by 2025, similar to the level in Japan today.
- Macroeconomic and labor-force implications:
  - Despite digitalization-led efficiency gains, China’s potential growth will slow as the economy matures; digitalization can moderate but not reverse the downward trend.
  - Shift from industry to services since 2012 expected to continue, putting downward pressure on potential growth because services productivity is lower than industrial productivity in China.
  - Industrial job losses likely; overall employment impact likely contained if workers transition to services.
  - Expected slowdown in employment growth rate to below 0.1 percent.
  - Digitalization may increase labor market polarization, “hollowing out” mid-skilled jobs while raising wages and employment for high- and low-skill labor.
  - Downside risk if AI is applied on a large scale: potential for larger-than-anticipated labor market disruption.

### Conclusions and policy recommendations
- Summary conclusion: Digitalization brings significant benefits (boosting productivity, promoting rebalancing, creating jobs in new sectors) but also risks (disruption of traditional sectors, job losses in mid-skill manufacturing, financial stability concerns from fintech).
- Recommended government actions:
  - Active labor market policy: retraining the labor force to smooth transitions.
  - Strengthen social safety nets for affected workers close to retirement to ensure minimum living standards.
  - Competition policy: promote competition, encourage data sharing, consider policies supporting open networks and foreign firm entry; note that local data storage and cross-border data transfer requirements may hinder global integration.
  - Comprehensive fintech supervisory oversight: coordinate regulators and industry to close loopholes, focus on “form over substance” regulation, use regulatory sandboxes, reflect fintech risks in liquidity buffers and capital adequacy, strengthen data gathering and KYC for third-party payments.
  - Data privacy and consumer protection: supervisors coordinate with relevant authorities to enhance regulation; establish laws to protect consumer data privacy and jurisdictions over data sharing.
  - E-government: improve digitalization of public services; current UN e-government index rank for China in 2016: 63rd out of 193 countries.
  - Continued support for digital infrastructure and internet penetration improvements, especially in rural areas.
  - Education policy: increase focus on ICT education in universities and vocational schools.
  - More active role in setting global standards on regulation, cyber security, global digital standards and governance.

*Italic: Source: wp1916 - 2.05 million tons and the cumulative trees planted amounting to 13.14 million.*

### References:

### wp1916 - References

### References
- AliResearch, 2016, Three Future Trends in the E-commerce Logistics Industry, May 2016.
- AliResearch, 2017A, Five Years of Innovation-10 Key Words about China`s Internet Industry, October 2017.
- AliResearch, 2017B, Digital Economy 2.0, January 2017.
- AliResearch, 2017C, Report on China’s Taobao Village, December 2017.
- Basel Committee on Banking Supervision, 2018, Sound Practices – Implications of Fintech Development for Bank and Bank Supervisors, February 2018.
- Boston Consulting Group, 2017 A, Towards 2035: The Future with 400 billion Digital Employment, , January 2017.
- Boston Consulting Group 2017 B, Towards 2035: The Battle on Human Capital for Digital Economy, January 2017.
- China Academy of Information and Communications Technology, 2017A, White paper on digital economy development, July 2017.
- China Academy of Information and Communications Technology, 2017B, White paper on digital economy development of G20 countries, December 2017.
- China Academy of Information and Communications Technology, 2017C, Internet Development Trends Report  , December 2017.
- China Development Forum, 2017, The Future of Artificial Intelligence in China, March 2017.
- China Internet Information Center, 2017, China statistical report on internet development, January 2017
- Das, M. and B. Hilgenstock, 2018, “The Exposure to Routinization: Labor Market Implications for Developed and Developing Economies”, IMF working paper 18/135.
- Didi Taxi 2017, The Big Data Report on Smart Travel, January 2017.
- Financial Stability Board and Committee on the Global Financial System (2017): FinTech credit: Market structure, business models and financial stability implications, May 2017
- Goldman Sachs, 2017, The Rise of China FinTech, August 2017
- Guizhou Global Big Data Exchange, White paper on China big data exchange, May 2016.
- Mckinsey, 2017A, Digital China: powering the economy to global competitiveness, December 2017.
- Mckinsey, 2016, Where machines could replace humans and where they can`t(yet), July 2016.
- Mckinsey, 2017B, Artificial Intelligence: The next digital frontier? June 2017.
- Tencent Research Institution, 2017, The 2017 China “Internet Plus” digital economy index, April 2017.
- U.S. Department of Commerce, The Emerging Digital Economy, 1998, 1999.
- World Bank, 2016, World Development Report 2016: Digital dividends, January 2016.
- World Bank and the People’s Bank of China, 2018, Toward Universal Financial Inclusion in China, February 2018
- World Economic Forum, 2016, The Global Information Technology Report 2016, July 2016.
- Zhang, L., 2016, “Rebalancing in China—Progress and Prospects,” IMF working paper 16/183.

### Technical Appendix: Impact of Digitalization on Productivity and Employment

- Scope:
  - Quantifies the impact of digitalization on productivity and employment using regression analysis at the national level and panel regression at the provincial level.

- Least square regression at national level:
  - Data:
    - annual data on digital economy from China Academy of telecommunication Research, interpolated into quarterly data (DEI).
    - GDP, investment (INV), and employment (EMP) are quarterly data from National bureau of statistics.
    - All variables are expressed in growth terms.
  - Regression equations (as presented):
    - 0123
      GDPDEI
      ttlttt
      INVEMPααα   α    ε
      −
      =+×   +    ++
    - 012
      DEI
      ttltt
      EMPGDPαααε
      −
      =+×   ++
  - Results: Impact on growth (table reproduced as in source)
    - Models: Model 1 | Model 2 | Model 3
    - VARIABLES — GDP growth | GDP growth | GDP growth
    - lagged digitalization: 0.34* | 0.47*** | 0.29
      - (0.18) | (0.12) | (0.26)
    - investment: 0.68*** | 0.71***
      - (0.11) | (0.12)
    - labor: 7.9 | 24.6***
      - (8.1) | (0.03)
    - Constant: -1.8 | -3.87
      - (1.98) | (2.01)
    - Observations: 444444
    - R-squared: 0.56 | 0.55 | 0.31
    - Robust standard errors in parentheses
    - Significance notation: *** p<0.01, ** p<0.05, * p<0.1

- Impact on employment:
  - Panel regression at provincial level:
    - Data:
      - annual data on digital economy index from TenCent, interpolated into quarterly data (DEI).
      - Provincial GDP, investment (INV), are quarterly data from National bureau of statistics.
      - All variables are expressed in log terms, except GDP/Capita in level term.
    - Regression equation (as presented):
      - ,01,3,12,
        GDPDEI/
        ititittitit
        GDP   CapitaINVαααα   ηχε
        −
        = +×   +×+    +++
  - Results: Employment Growth (panel)
    - lagged digitalization: 0.013***
      - (0.002)
    - GDP growth: 0.005***
      - (0.002)
    - Constant: 0.007
      - (0.04)
    - Observations: 49
    - R-squared: 0.57
    - Robust standard errors in parentheses
    - Significance notation: *** p<0.01, ** p<0.05, * p<0.1

  - Results: Full Sample / Developed Regions / less developed regions (GDP growth regressions; table reproduced)
    - VARIABLES — GDP growth | GDP growth | GDP growth
    - lagged digitalization: 0.27*** | 0.35*** | 0.27**
      - (0.05) | (0.04) | (0.09)
    - investment: -0.05 | 0.13 | -0.05
      - (0.03) | (0.1) | -0.04
    - GDP/Capita: -4.6 | -4.25*** | -6.5 ***
      - (1.14) | (0.9) | (2.94)
    - Constant: 12.5*** | 12.8*** | 12.7***
      - (1.00) | (1.10) | (2.14)
    - Observations: 146 | 68 | 78
    - Provinces: 30 | 14 | 16
    - R-squared within: 0.26 | 0.53 | 0.22
    - R-squared between: 0.12 | 0.01 | 0.03
    - R-squared overall: 0.11 | 0.01 | 0.03
    - Robust standard errors in parentheses
    - Significance notation: *** p<0.01, ** p<0.05, * p<0.1

*Source: wp1916 - References (technical appendix and bibliography).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wp1916.pdf_