## sdnea2023003 - Executive Summary

## Source details

**Canonical URL:** [sdnea2023003 - Executive Summary](https://www.imf.org/-/media/files/publications/sdn/2023/english/sdnea2023003.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/sdn/2023/english/sdnea2023003.pdf.md)
- [Structured JSON version](/-/media/files/publications/sdn/2023/english/sdnea2023003.pdf.json)

---

### Scope and caution
- Focus: digitalization during the COVID-19 crisis and implications for productivity and labor markets in selected advanced economies.
- Time coverage: analysis based on two full years of data; results are preliminary and require caution in interpreting permanence.

### Acceleration of digitalization
- Primary measure: share of workers using a computer with access to the internet (Eurostat).
- Aggregate changes:
  - "the share of workers using a computer with internet access increased from 56 to 61 percent on average, about a 10 percent increase."
  - "on average, digitalization increased above trend during the pandemic from 56 percent of workers using a computer with internet access to 62 percent, a 6 percentage point rise (10 percent in proportional terms)."
- Cross-country and cross-sector patterns:
  - Sweden: 82 percent of workers used a computer with internet access in 2019.
  - Greece: 38 percent of workers used a computer with internet access in 2019.
  - Countries with lower initial digitalization experienced the largest increases; Greece experienced an increase of almost 20 percent.
  - Contact-intensive sectors (food and accommodation, construction) and smaller firms showed the largest increases in digitalization.
- US evidence (software spending per employee proxy):
  - Software spending per employee increased in 2020 across all industries relative to the latest pre-COVID year.
  - Relative largest rises in historically lower-digitalized sectors (arts and entertainment; food and accommodation).
  - Part of the increase reversed in 2021 but overall spending remained higher than pre-COVID levels except in ICT and health and social sectors.

### Effects on productivity
- Sectoral labor productivity (sector-level analysis):
  - At the height of the pandemic (2020), higher sectoral digitalization reduced labor productivity losses by 20 percent when comparing the 75th and 25th percentiles of digitalization.
  - If countries with low digitalization had had the 75th percentile digitalization in each sector, aggregate labor productivity growth during the pandemic would have been higher by a quarter.
  - Counterfactual: average growth rate of aggregate productivity in 2020 in these countries would have been higher by 0.6 percentage point—approximately a quarter of their effective growth rate.
  - Heterogeneity:
    - For non-contact-intensive sectors, the contraction for a highly digitalized sector (75th percentile) is 40 percent smaller than for one with low digitalization (25th percentile) in 2020.
    - Even in contact-intensive sectors, higher digitalization helped shield labor productivity in 2020.
- Firm-level total factor productivity (TFP):
  - Pandemic accelerated a slowdown of TFP that started in 2018, lowering it by 1.8 percent in 2020 relative to 2019.
  - In 2021, TFP growth on average reversed, catching up to and—in the US—almost reaching pre-pandemic levels.
  - Highly digitalized firms’ TFP recovered more strongly from the pandemic than that of low-digitalized firms (US and advanced Europe).
  - Caveat: larger firms may be better positioned to leverage digital investments due to economies of scale and network effects.
- Regression and causal notes:
  - US firm-level regression (2011–19): a 10 percent increase in gross computer equipment and software assets per worker is associated with:
    - a 0.03 percentage point increase in TFP growth the following year; and
    - a cumulative 0.3 percentage point increase in long-term TFP.
  - Results show implementation lags; contemporaneous coefficients often not statistically significant.
  - Results should not be interpreted as causal without caveats; two-stage least squares with sector-wide instruments mitigates endogeneity concerns.

### Effects on employment and hours worked
- Hours worked and sectoral protection:
  - For a high level of sectoral digitalization (75th percentile) relative to a low level (25th percentile):
    - sector-level losses in hours worked would have been 85 percent smaller among non-contact-intensive sectors in 2020; and
    - sector-level losses in hours worked would have been 40 percent smaller in contact-intensive sectors in 2020.
  - Aggregate counterfactual: countries with low 2019 digitalization could have reduced losses in hours worked by slightly more than one-third of the total contraction if their sectors had digitalization levels equal to the 75th percentile observed within each sector in the sample.
  - Statistical note: the interaction coefficient is significant at the 10 percent level and when country-level population weight is used, suggesting results driven by larger European countries.
  - In many European economies, job retention programs caused most contraction in total hours worked to occur along the intensive margin (hours per worker) rather than the extensive margin (number of workers).

### Occupation- and skill-level impacts
- Digital occupations:
  - Definition: occupations above the 50th percentile of O*NET digital scores based on knowledge and frequency of computer use.
  - Employment share of digital occupations increased over time and accelerated during the pandemic.
  - Countries with lower initial 2015 digital employment share saw catch-up during 2019–22.
  - Examples:
    - Portugal saw the fastest acceleration in digital employment share during the pandemic.
    - Denmark did not see a sustained increase; Finland saw a decline relative to trend.
- Protection and demand:
  - Workers in digital occupations were less affected by layoffs than workers in non-digital occupations, particularly in countries without job retention programs.
  - Little evidence so far of a structural shift in labor demand toward digital occupations; vacancies data show an initial increase in the share of digital occupations that subsequently reversed.
  - Labor markets tightened more for low-skilled jobs than for higher-skilled ones during the pandemic.

### Vacancies and labor demand for digital occupations
- Vacancies data (Indeed norm-cats and micro postings):
  - The share of vacancies for digital occupations increased only slightly since 2020 in the average country.
  - Heterogeneity across countries:
    - Spain: digital vacancy share rose by 5 percentage points by 2022.
    - Luxembourg: digital vacancy share fell by about 5 percentage points.
    - France, the UK, Italy, Ireland: digital vacancy share declined.
  - About half of sample countries saw only a temporary increase in digital vacancies in 2020 and a reversal in 2021–22.
- Regional regression evidence (US and UK):
  - States/regions more exposed to COVID-19 saw a larger and more persistent rise in the digital vacancies share through mid-2021; by 2022 the differential impact waned.
  - Cross-country analysis: larger exposure to COVID-19 associated with a statistically significant temporary rise in digital vacancies during Q2 2020 only.
- Additional channels sustaining demand for non-digital occupations:
  - Increased inactivity among older and less educated workers created shortages in non-digital occupations.
  - Decline in net inflows of migrant workers aggravated shortages.
  - Consumption swings increased demand for manual-intensive jobs (warehousing, food and package delivery).
- Intensive margin of digitalization:
  - No signs that the pandemic increased within-occupation demand for digital skills (LinkedIn data show no increase in the intensive margin).

### Working from home, labor supply, and time use
- Telework rates in Europe:
  - Before COVID-19: 5 percent usually worked from home.
  - 2020: share increased to 10 percent.
  - 2021: share increased to 16 percent.
  - Cross-country 2021 examples: Greece 7 percent; Finland 25 percent.
- Gap and preferences:
  - Gap between potential (share in teleworkable sectors) and actual work-from-home rates declined but remains.
  - Estimated amenity gains from working from home range from 1.5 percent of earnings at the low end to 7.3 percent at the high end.
  - These amenity gains may help explain muted wage growth after the pandemic in the US despite labor market tightness.
- Associations with labor force participation and transitions:
  - Countries with larger shares working from home experienced smaller drops or increases in labor force participation relative to trend in 2021.
  - UK: before the pandemic teleworkers were more likely to leave the labor force (especially females); by 2022 this gap shrank.
  - US: transition rates into inactivity were lower for workers who had worked from home due to COVID-19.
- Time-use evidence:
  - Teleworkable occupations save on average two hours a week by not commuting (ATUS).
  - Survey evidence finds workers allocate 40 percent of commuting time savings to their jobs.
  - Saved commuting time associated with increased working hours for women in teleworkable occupations and a decline in working hours for men.
- Working-from-home model (Annex 6) findings:
  - Decline in commuting time increases labor supply and leisure for teleworkable workers.
  - Increase in men's preference for leisure reduces men's labor supply and raises women's labor supply via family income effects.
  - Combined shocks replicate observed patterns: increase in women's working hours and decline in men's working hours.
  - Conclusion: increase in remote work improves welfare and is associated with an increase in labor supply of workers in teleworkable occupations.

### Nature of the digitalization observed
- Observed increase concentrated in "primary digitalization" enabling operations without in-person interactions (remote operations, online sales, contactless payments) rather than broad adoption of advanced technologies (AI, cloud, big data).
- Advanced technologies:
  - Except for cloud computing, measures of advanced technologies have low penetration in most sectors, smaller cross-country and cross-sector differences, and were less dynamic during COVID-19.

### Digital gaps, connectivity costs, and skills
- Connectivity costs and digital gaps:
  - Cost of fixed-broadband internet in Greece and Portugal is almost two times that of Norway; internet speed is slower in less digitalized countries.
  - Less digitalized countries closed 20 percent of the gap with frontier-digitalization countries during the pandemic, but large gaps remain even after controlling for sectoral composition.
- Human capital and STEM:
  - Among workers in highly digital occupations, 74 percent have a STEM degree, including master's and doctoral degrees.
  - In Portugal fewer than 8 percent of workers have postgraduate STEM education; in Sweden almost 15 percent have postgraduate STEM education.
  - Moving up the digital ladder requires increasing the share of workers with digital skills.

### Policy recommendations
- Priorities to close digitalization gaps and ensure broad-based gains:
  - Invest in digital infrastructure.
  - Help workers acquire needed skills, including STEM skills.
  - Maintain healthy competition in digital markets.
  - Adapt labor laws and regulations to changing work arrangements to facilitate telework or remote work access.
- Further policy specifics:
  - Bring down the cost of internet to support adoption by small firms and households.
  - Ensure competition in telecommunications to reduce costs and increase access to digital technologies.
  - Balance incentives to invest in productive digitalization versus stimulating excessive digitalization that accelerates substitution of labor for capital.
  - Adapt competition policy to digital markets (data portability, interoperability, fair and transparent practices).
  - Invest in human capital (education and training) to equip the workforce with skills for higher-tech tasks.
  - Adapt labor laws to keep workers accessible to firms while protecting work-life balance and privacy; share costs of resources used to work from home.
- NGEU / RRF note (Annex 7):
  - Proposed RRF digital support ranges from approximately 1.3 percent of 2020 GDP (Greece) to 0.08 percent of 2020 GDP (Ireland).
  - Less digitalized countries proposed more ambitious RRF digital allocations, consistent with catch-up objectives.

### Conclusions and overarching findings
- The COVID-19 pandemic induced a digitalization push concentrated in contact-intensive sectors and among small firms that started with relatively low levels of digitalization.
- Evidence points to catching up in the use of essential technologies rather than increased investment in advanced technologies.
- Digitalization enhances resilience and could increase productivity in the longer term:
  - Higher levels of digitalization were associated at the sector level with smaller losses in productivity and especially in employment in response to the pandemic shock.
  - Differences in digitalization across countries can explain a non-negligible part of economic underperformance of less digitalized economies during the pandemic.
  - Historical evidence suggests digitalization could lead to gains in total factor productivity over the longer term; benefits for smaller firms depend on policies ensuring healthy competition in digital markets.
- Work-from-home changes appear more persistent so far and could boost labor supply by reducing disutility associated with work, increasing labor market attachment, and improving workers’ welfare.
- Governments need to invest in digital infrastructure and education and adapt competition and labor policies to ensure the gains from digitalization are broadly shared.

*International Monetary Fund — Staff Discussion Note: Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies*

### Executive Summary ......................................................................................................

### sdnea2023003 - Executive Summary

### I. Introduction
- Contents listed on page 5.
- Sets up the study: "Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies."

### II. Stylized Facts
- Contents listed on page 7.
- Presents stylized facts underpinning the analysis.

### III. Productivity
- Contents listed on page 10.
- Subsections:
  - III.1 Digitalization and Labor Productivity (page 11)
  - III.2 Digitalization and Firm Total Factor Productivity (page 13)
- Figures related to productivity:
  - Figure 1. Labor Productivity and Labor Market Tightness (page 5)
  - Figure 3. Digital Intensity Increased in the US (page 10)
  - Figure 4. Covid-19 Has Impacted Negatively Labor Productivity (page 11)
  - Figure 5. Regression Results on Labor Productivity (page 12)
  - Figure 6. Evolution of Firms’ Total Factor Productivity (page 14)
  - Figure 7. Firms’ ICT Intensity Has Increased Since 2019 (page 15)
- Annexes linked to productivity analysis:
  - Annex 2. Digital Intensity Decomposition (page 38)
  - Annex 3. Productivity Regression Analysis (page 39)

### IV. Labor Markets
- Contents listed on page 15.
- Subsections:
  - IV.1 Digitalization and Sectoral Employment (page 16)
  - IV.2 Digital Occupations (page 17)
  - IV.3 The Impact of COVID-19 on Digital Labor Demand (page 19)
  - IV.4 The Effects of Digitalization on Labor Supply (page 21)
- Figures related to labor markets:
  - Figure 8. Regression Results on Total Hours Worked (page 16)
  - Figure 9. Digital Occupations (page 17)
  - Figure 10. Regression Analysis of the Effect of Covid-19 on Employment in Digital an non-Digital Occupations (page 18)
  - Figure 11. Share of Vacancies in Digital Occupations in Advanced Economies (page 19)
  - Figure 12. Regression Analysis of Covid-19 Exposure and Vacancies for Digital Occupations (page 20)
  - Figure 13. Trends in Working from Home (page 21)
  - Figure 14. Working from Home and Labor Force Attachment (page 22)
  - Figure 15. The Labor Supply Implications of Working from Home (page 23)
- Annexes linked to labor market analysis:
  - Annex 4. Employment Regression Analysis (page 46)
  - Annex 5. Vacancies Regression Analysis (page 48)
  - Annex 6. Working From Home (page 49)

### V. Policy Recommendations
- Contents listed on page 24.
- Figure related to policy role:
  - Figure 16. The Role of Policies on Digitalization (page 24)

### VI. Conclusion
- Contents listed on page 25.

### Boxes
- Box 1. Alternative Measures of Digitalization (page 27)
- Box 2. Measuring the Evolution of Digital Skills Using Data from LinkedIn (page 28)

### Figures (Inventory)
- Figure 1. Labor Productivity and Labor Market Tightness
- Figure 2. Covid-19 Has Accelerated Digitalization in Advanced European Countries
- Figure 3. Digital Intensity Increased in the US
- Figure 4. Covid-19 Has Impacted Negatively Labor Productivity
- Figure 5. Regression Results on Labor Productivity
- Figure 6. Evolution of Firms’ Total Factor Productivity
- Figure 7. Firms’ ICT Intensity Has Increased Since 2019
- Figure 8. Regression Results on Total Hours Worked
- Figure 9. Digital Occupations
- Figure 10. Regression Analysis of the Effect of Covid-19 on Employment in Digital an non-Digital Occupations
- Figure 11. Share of Vacancies in Digital Occupations in Advanced Economies
- Figure 12. Regression Analysis of Covid-19 Exposure and Vacancies for Digital Occupations
- Figure 13. Trends in Working from Home
- Figure 14. Working from Home and Labor Force Attachment
- Figure 15. The Labor Supply Implications of Working from Home
- Figure 16. The Role of Policies on Digitalization

### Annexes and Reference
- Annex 1. Data Sources, Sample Coverage, and Variable Definitions (page 29)
- Annex 2. Digital Intensity Decomposition (page 38)
- Annex 3. Productivity Regression Analysis (page 39)
- Annex 4. Employment Regression Analysis (page 46)
- Annex 5. Vacancies Regression Analysis (page 48)
- Annex 6. Working From Home (page 49)
- Annex 7. Box A1 (page 51)
- Reference section (page 52)

*Staff Discussion Note: Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies*

### Executive Summary

### Executive Summary

### Scope and caution
- Focus: digitalization during the COVID-19 crisis and implications for productivity and labor markets in selected advanced economies.
- Time coverage: analysis based on two full years of data; results are preliminary and require caution in interpreting permanence.

### Acceleration of digitalization
- Primary measure: share of workers using a computer with access to the internet (Eurostat).
- Aggregate changes reported in the source:
  - "the share of workers using a computer with internet access increased from 56 to 61 percent on average, about a 10 percent increase."
  - Elsewhere: "on average, digitalization increased above trend during the pandemic from 56 percent of workers using a computer with internet access to 62 percent, a 6 percentage point rise (10 percent in proportional terms)."
- Cross-country and cross-sector patterns:
  - Sweden: 82 percent of workers used a computer with internet access in 2019.
  - Greece: 38 percent of workers used a computer with internet access in 2019.
  - Countries with lower initial digitalization (including southern and eastern European countries) experienced the largest increases; Greece experienced an increase of almost 20 percent.
  - Contact-intensive sectors (food and accommodation, construction) and smaller firms showed the largest increases in digitalization.
- US-specific evidence:
  - Digitalization proxied by software spending per employee increased in 2020 relative to the latest pre-COVID year across all industries.
  - In relative terms, some of the largest rises occurred in sectors with historically lower digitalization such as arts and entertainment and food and accommodation.
  - Part of the increase in digital spending reversed in 2021 but overall spending remained higher than pre-COVID levels except in ICT and health and social sectors.

### Effects on productivity
- Pandemic-period shielding:
  - At the height of the pandemic (2020), higher sectoral digitalization reduced labor productivity losses by 20 percent when comparing the 75th and 25th percentiles of digitalization.
  - If countries with low digitalization had had the 75th percentile digitalization in each sector, aggregate labor productivity growth during the pandemic would have been higher by a quarter.
- Subsequent rebound dynamics:
  - In 2021, sector-level labor productivity experienced a partial rebound that was larger in less digitalized sectors, which had experienced larger declines in 2020.
- Firm-level TFP evidence:
  - For publicly listed firms in the US and advanced European economies, highly digitalized firms’ TFP recovered more strongly from the pandemic than that of low-digitalized firms.
  - Caveat: stronger TFP gains may reflect that larger firms are better positioned to leverage digital investments due to economies of scale and network effects.

### Effects on employment and hours worked
- Hours worked:
  - At the sectoral level, greater digitalization translated into smaller losses in hours worked, especially in non-contact-intensive sectors, reducing these losses by 85 percent between the 75th and 25th percentiles of digitalization.
  - At the country level, if less digitalized countries had had a digitalization level equivalent to the 75th percentile of each sector, employment losses from pandemic-induced disruptions would have been a third lower in 2020.
- Occupation- and skill-level impacts:
  - Workers in digital occupations were less affected by layoffs than workers in non-digital occupations, particularly in countries without job retention programs; effect broad-based across digital occupations.
  - Despite better protection for digital occupations during the crisis, there is little evidence so far of a structural shift in labor demand toward digital occupations; vacancies data show an initial increase in the share of digital occupations that subsequently reversed.
  - Labor markets tightened more for low-skilled jobs than for higher-skilled ones during the pandemic.

### Nature of the digitalization observed
- The increase in digitalization appears concentrated in "primary digitalization" that enabled firms to operate without in-person interactions (remote operations, online sales, contactless payments) rather than widespread adoption of advanced technologies (AI, cloud, big data).
- Advanced technologies:
  - Except for cloud computing, measures of advanced technologies have low penetration in most sectors, smaller cross-country and cross-sector differences, and were less dynamic during the COVID-19 period.

### Working from home and labor supply
- Telework transformation:
  - Before the crisis in Europe, on average only 5 percent of workers typically worked from home.
  - By 2021, the share of workers who usually work from home increased to more than 16 percent.
  - Working from home increases worker welfare by reducing commuting time and improving time management flexibility and could boost labor supply by reducing the disutility of work.

### Distributional and structural considerations
- Preexisting gaps:
  - Substantial differences in digitalization across countries, sectors, and firms prior to the crisis; small firms were substantially less digitalized than medium and large firms.
  - Differences across countries were mainly due to within-sector digitalization gaps rather than sectoral composition; linked to the cost of internet broadband (e.g., cost of fixed-broadband internet in Greece and Portugal almost two times that of Norway).
- Potential for unequal gains:
  - Larger gains in productivity and TFP have been observed for larger, more digitalized firms; whether smaller firms can reap similar benefits may depend on policies to ensure healthy competition and to address economies of scale and network effects.

### Policy recommendations
- Priorities to close digitalization gaps and ensure broad-based gains:
  - Invest in digital infrastructure.
  - Help workers acquire needed skills, including STEM skills.
  - Maintain healthy competition in digital markets.
  - Adapt labor laws and regulations to changing work arrangements to facilitate telework or remote work access.
- Balance objectives:
  - Governments should balance supporting productive firm-level digitalization and equipping workers to adapt to a more digital economy.
  - Digitalization can increase economic resilience to shocks and encourage greater labor force participation through flexible work arrangements.

*International Monetary Fund — Staff Discussion Note: Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies*

### Annex Figure 1.1 shows that results hold when using the share of software spending in total current expenditures.

### sdnea2023003 - Annex Figure 1.1 shows that results hold when using the share of software spending in total current expenditures.

### Productivity: channels and measurement
- Multiple channels through which digitalization may have increased productivity:
  - Purchases of equipment and software platforms (examples: Zoom and Microsoft Teams) to enable working from home and facilitate meetings and scheduling.
  - Savings in commuting time, reduced attrition, facilitated interaction, and enhanced job satisfaction (Bloom, Han, and Liang 2022).
  - Investments in digital platforms to improve logistics and online sales, extending market reach and operations.
- Historical literature finds IT had a positive impact on productivity and growth (Brynjolfsson and Yang 1996; Brynjolfsson and Hitt 2000; Oliner and Sichel 2000; Crépon and Heckel 2002).
- Two measures of productivity used in the analysis:
  - Labor productivity: output per hour worked — advantage: covers entire economy or sector; drawback: impacted by changes in capital investment or skill composition.
  - Firm-level total factor productivity (TFP): cleaner measure of underlying productivity but available only for a sample of publicly listed firms.
- Note: Focusing on output per hour worked rather than output per worker helps account for diverse coverage of job retention programs across countries and sectors.

### Digitalization and labor productivity (sectoral analysis)
- Rationale:
  - Aggregate labor productivity increased during the pandemic partly due to labor shedding and reallocation toward more productive sectors; therefore sectoral analysis helps isolate digitalization effects beyond cyclical behavior.
  - The analysis uses a Google mobility indicator to capture severity of the pandemic and lockdown measures and exploits differences in digitalization across sectors and countries.
- Heterogeneous sectoral and country impacts:
  - Industries with lower levels of digitalization in 2019 (examples: food and accommodation; transport and storage) generally experienced deeper productivity losses over 2020–21, though there is marked within-sector heterogeneity across countries.
  - Ireland highlighted for notable labor productivity growth, tied to ICT hub status.
- Regression results (panel of nine sectors in 22 advanced European economies, 2020–21):
  - In 2020 the sectoral productivity loss in response to the pandemic shock was 20 percent smaller for a high level of digital intensity (75th percentile of the sector-adjusted distribution) than for a low level of digital intensity (25th percentile).
  - In 2021 less digitalized sectors rebounded more strongly than more digitalized sectors (partial catch-up), so cumulatively over 2020–21 higher digitalization was associated with a modest buffering of the productivity drop.
- Contact-intensity differences:
  - When allowing the mobility–productivity association to differ for contact-intensive and non-contact-intensive sectors:
    - For non-contact-intensive sectors, the contraction for a highly digitalized sector (75th percentile) is 40 percent smaller than for one with low digitalization (25th percentile) in 2020.
    - The association between digitalization and productivity is more pronounced in non-contact-intensive sectors for two reasons: higher responsiveness of productivity to digitalization, and larger dispersion of digital intensity across countries in these industries.
  - Even in contact-intensive sectors, higher digitalization helped shield labor productivity in 2020.
- Aggregate counterfactual exercise:
  - Constructing counterfactual country-level productivity series (assuming low-digitalization countries had instead the average or 75th percentile level of digitalization observed in the sample for each sector) implies:
    - Sector-level labor productivity losses associated with the Google mobility shock would have been reduced.
    - The average growth rate of aggregate productivity in 2020 in these countries would have been higher by 0.6 percentage point—approximately a quarter of their effective growth rate.

### Digitalization and firm total factor productivity (TFP)
- TFP dynamics for publicly listed firms (US and advanced Europe):
  - The pandemic accelerated a slowdown of TFP that started in 2018, lowering it by 1.8 percent in 2020 relative to 2019.
  - In 2021 there was on average a reversal, with TFP growth catching up to and—in the US—almost reaching the pre-pandemic level.
  - The recovery of TFP growth was more anemic in advanced European economies compared with the US; large cross-country heterogeneity in pre-pandemic digitalization corresponds with large differences in TFP growth recovery.
- Firm-level digital intensity and recovery:
  - Highly digitalized firms definition:
    - US: firms in the top 25 percent of the distribution of the share of computer and software assets per worker.
    - Advanced Europe: firms in country-sector cells in the top 25 percent of the distribution of the share of workers with internet access.
  - By 2021, using a balanced panel of firms, the gap between the TFP growth rate of more and less digitalized firms was growing (more digitalized firms recovered faster).
- Regression evidence on within-firm TFP growth:
  - US analysis (controls include state-industry fixed effects and firm characteristics such as age, pre-pandemic size, profitability) indicates that firms more digitalized ex ante experienced relatively larger within-firm TFP growth in 2021 compared with less digitalized firms.
  - For 2020, regression results suggest that the impact of ICT preexposure was not statistically different from zero in the US.
  - Comparable regression analysis for advanced Europe yields similar results: ICT pre-adoption helped firm productivity recover faster from the crisis in 2021.
- Interpretation:
  - TFP results for publicly listed firms provide stronger evidence of lasting benefits from digitalization than sectoral labor productivity evidence over the 2020–21 span.
  - Differences between labor productivity and firms’ TFP results may reflect factors including sample (publicly listed, larger, more digitalized firms) and measurement differences.

*Source: STAFF DISCUSSION NOTES — Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies (INTERNATIONAL MONETARY FUND).*

### Annex 3.2 provides further details on the regression analysis. For the US, variation of pre-pandemic ICT intensity at th

### 19. The cross-country analysis suggests that a larger exposure to COVID-19 was associated with a

### 19. The cross-country analysis suggests that a larger exposure to COVID-19 was associated with a

### Impact on vacancies and occupation-level demand
- Larger exposure to COVID-19 was associated with a statistically significant temporary rise in the differential growth rates of vacancies for digital occupations during the second quarter of 2020 only (Figure 12, panel 2).
- Exploiting cross-regional variation in the US and the UK yields a larger and more persistent impact of exposure to the pandemic, with the positive effect on the digital vacancies share on more exposed regions persisting through mid-2021 (Figure 12, panels 1).
- By 2022 the differential impact had waned within both countries.
- Multiple channels explain resilience of labor demand for non-digital occupations:
  - Increased inactivity, particularly among older and less educated workers, resulted in labor shortages in industries and occupations where these demographic groups are employed—mostly non-digital occupations; vacancies in these jobs grew rapidly as firms tried to cover the workforce shortfall.
  - Decline in net inflows of migrant workers in some countries further aggravated shortages.
  - Abrupt swings in consumption patterns during lockdowns shifted spending away from some in-person services toward goods purchasable remotely and consumed at home, increasing demand for manual-intensive jobs such as warehousing and food and package delivery.
- There are no signs that the pandemic increased the "intensive margin" of digital employment by increasing the demand for digital skills within occupations. Analysis using LinkedIn data finds no evidence of an increase in the intensive margin of digitalization during the crisis.
- Supporting citations from related studies in the text:
  - Bellatin and Galassi (2022) using Indeed data for Canada show digital-related job postings fared relatively better when policies became more stringent.
  - Copestake and others (2022) find digitalization helped shield workers in Asia and EMDEs.
  - Alcedo and others (2023) find a significant transient increase in online spending as a share of total consumption during the pandemic; effects dissipate over time.
  - Forsythe and others (2022) found evidence of down-skilling after the pandemic.

### Work from home, labor supply, and time use
- The COVID-19 crisis led to a significant and persistent increase in rates of work from home:
  - Before COVID-19, only 5 percent of workers usually worked from home in Europe.
  - In 2020 the share increased to 10 percent.
  - In 2021 the share increased to 16 percent.
- Cross-country differences in 2021:
  - Greece: share of workers teleworking increased to 7 percent in 2021.
  - Finland: share increased to 25 percent in 2021.
- The gap between potential (share of workers in teleworkable sectors) and actual rates of work from home declined significantly following the COVID-19 crisis, but a significant gap remains.
- Workers highly value working from home; estimated amenity gains range from 1.5 percent of earnings at the low end of the earnings distribution to 7.3 percent at the high end (Barrero, Bloom, and Davis 2021).
- These amenity gains may help explain muted wage growth after the pandemic in the US despite labor market tightness (Barrero and others 2022); lack of wage growth is especially notable for high-skilled workers, while low-skilled workers experienced wage gains, contributing to a recent decline in wage dispersion (Autor and Dube 2022).
- Associations with labor force participation:
  - Countries with a larger share of workers working from home experienced a smaller drop in labor force participation and even an increase in labor force participation relative to trend in 2021 (Figure 14, panel 1).
  - The stronger association in 2021 than 2020 suggests working from home may attract marginally attached workers and extend the working life of elderly workers.
- Transition rates to inactivity (UK and US evidence):
  - In the UK, before the pandemic those working from home were more likely to leave the labor force than on-site workers, particularly females; by 2022 this gap shrank, suggesting greater labor force attachment among those working from home.
  - In the US, transition rates into inactivity were lower for both male and female workers who had previously worked from home due to COVID-19.
- Time-use evidence:
  - Workers in teleworkable occupations save on average two hours a week by not commuting to an office (ATUS).
  - Survey evidence (Aksoy and others 2023) finds workers allocate 40 percent of their commuting time savings to their jobs.
  - Saved commuting time is associated with an increase in working hours for women in teleworkable occupations, but a decline in working hours for men in teleworkable and non-teleworkable occupations.
  - The reduction in men's working hours suggests an increased preference for leisure for men, potentially accelerating a long-term trend in men's decline in labor market attachment.

### Modeling of labor supply effects
- A household labor supply model (Annex 6) simulates four household types (husband and wife working in teleworkable or non-teleworkable occupations) calibrated with ATUS data and considers two shocks:
  1. Decline in commuting time impacting workers in teleworkable occupations.
  2. Increase in the preference for leisure affecting men only.
- Main model findings:
  - Decline in commuting time associated with remote work increases labor supply and leisure for workers in teleworkable occupations.
  - Increase in men's leisure preferences leads to a decline in men's labor supply and an increase in women's labor supply due to the decline in family income.
  - Combined effects replicate observed patterns: increase in women's working hours and decline in men's working hours.
  - Conclusion: increase in remote work improves welfare and is associated with an increase in labor supply of workers in teleworkable occupations.

### Digitalization, skills, and cross-country digital gaps
- Costs of fixed and mobile broadband internet vary across countries; countries with high internet costs have lower levels of digitalization. Example: cost of fixed-broadband internet in Greece and Portugal is almost twice as high as in Norway, and internet speed is slower.
- Less digitalized countries closed 20 percent of the gap with frontier-digitalization countries during the pandemic, but large gaps remain even after controlling for sectoral composition.
- Human capital and STEM education:
  - Among workers in highly digital occupations, 74 percent have a STEM degree, including master's and doctoral degrees.
  - More digitalized countries have a larger share of workers with a STEM degree, particularly with master's and doctoral education.
  - Example comparisons: in Portugal fewer than 8 percent of workers have postgraduate education in STEM, whereas in Sweden the number is almost 15 percent.
- Policy implications for skills: moving up the digital ladder requires increasing the share of workers with digital skills.

### Policy recommendations
- Invest in digital infrastructure and foment competition in the ICT sector to:
  - Increase cybersecurity, access to the cloud, and data processing capabilities.
  - Provide good internet to all households and reduce the digital divide.
  - Support small firms’ adoption of basic and advanced technologies, particularly as firms and households face increases in other costs of living.
- Bring down the cost of internet to support adoption by small firms and households.
- Ensure competition in telecommunications to reduce costs and increase access to digital technologies for small firms and poor households.
- Balance incentives to invest in productive digitalization versus stimulating excessive digitalization that accelerates substitution of labor for capital (Acemoglu, Manera, and Restrepo 2020; Acemoglu 2023).
- Adapt competition policy to digital markets:
  - Facilitate data portability and interoperability to increase competition with established players.
  - Establish fair, transparent, and inclusive competition practices to avoid winner-takes-all dynamics.
- Invest in human capital (education and training) to equip the workforce with skills to perform higher-tech tasks so gains from digitalization are inclusive.
- Adapt labor laws and regulations to changing work environments to facilitate telework or remote work access while balancing objectives:
  - Keep workers accessible to firms while protecting work-life balance.
  - Protect privacy of workers while allowing firms to monitor productivity.
  - Share costs of resources used to work from home.

### Conclusions and overarching findings
- The COVID-19 pandemic induced a digitalization push concentrated in contact-intensive sectors and among small firms that started with relatively low levels of digitalization.
- Evidence points to catching up in the use of essential technologies rather than increased investment in advanced or revolutionary technologies.
- Digitalization enhances resilience and could increase productivity in the longer term:
  - Higher levels of digitalization were associated at the sector level with smaller losses in productivity and especially in employment in response to the pandemic shock.
  - Differences in digitalization across countries can explain a non-negligible part of economic underperformance of less digitalized economies during the pandemic.
  - Historical evidence suggests digitalization could lead to gains in total factor productivity over the longer term; benefits for smaller firms depend on policies ensuring healthy competition in digital markets.
- Work-from-home changes appear more persistent so far and could boost labor supply by reducing disutility associated with work, increasing labor market attachment, and improving workers’ welfare.
- Governments need to invest in digital infrastructure and education and adapt competition and labor policies to ensure the gains from digitalization are broadly shared.

*Staff Discussion Notes — Digitalization during the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies*

### Box 1. Alternative Measures of Digitalization

### Box 1. Alternative Measures of Digitalization

### Overview of alternative measures
- Baseline proxy used in the note: the share of workers using a PC with internet access.
- Alternative measures from Eurostat’s Digital Economy and Society database (downloaded via OECD Stat, Information and Communication Technology section) measure the share of firms that:
  - purchase cloud services,
  - employ ICT specialists,
  - conduct “big data” analysis,
  - use the Internet of Things (IoT),
  - use artificial intelligence (AI),
  - use 3D printing.
- Data availability for these alternative measures is more limited in terms of country and time coverage than the baseline share of workers using a PC with internet access.

### Comparative patterns across sectors and countries
- The baseline measure (share of workers using an internet-connected computer) shows a marked gradual rise from the sector with lowest intensity (food and accommodation) to the highest (ICT), with large variation across countries.
- Alternative firm-level measures generally exhibit low penetration in most sectors.
- While the ranking of intensity is generally consistent across indicators, most alternative measures do not show pronounced differentials from sector to sector compared with the baseline worker-level measure.

### Interpretation and reasons for differences
- The share of workers using a computer reflects a basic and widespread type of digital technology that is multifunctional and potentially accessible to many workers.
- Alternative measures refer to more advanced digital technologies with larger fixed costs and are mostly used in larger firms.
- Consequently, alternative firm-level measures are less appropriate for providing a wide-angle view of digital technology use across a sector or country.
- The share of workers using a computer can be considered an upper bound for the use of more advanced technologies within a sector: deploying cloud computing, big data analysis, or artificial intelligence is challenging and likely inefficient if a larger fraction of employees does not first use computers in their work.

### Data note
- The alternative measures summarized in Box Figure 1.1 compare distribution by sectors across advanced European economies for the latest available data prior to the pandemic (or the earliest available data for two indicators) against the baseline measure.

*Prepared by Carlo Pizzinelli.*

### 1.5 Vacancies Data

### 1.5 Vacancies Data

### Data sources and series types
- Vacancies data are provided by Indeed, a large online job advertisement aggregator operating in numerous countries.
- Two types of data series are used:
  - Micro data of individual job posts (used in the regional analysis for the US and the UK).
  - Time series of vacancies aggregated by standardized categories of jobs called “norm-cats” (used in cross-country regressions).

### Individual-level vacancies data (US and UK)
- Sample period: January 2019 until June 2022.
- Each posting contains: the position’s job title, the date the advertisement was first posted, and the location within the country.
- Assumption: one vacancy per posting (supported by prior work noting this replicates time-series variation from traditional sources).
- Occupation assignment:
  - A text-matching algorithm assigns an occupation from the four-digit version of the 2008 International Standard Classification of Occupations (ISCO-08) to each job title.
  - Through the four-digit ISCO-08 code, individual vacancies are classified as digital or non-digital occupations using the O*NET digital score and the country-specific thresholds discussed in the main text.
- Aggregation:
  - Individual job postings are aggregated monthly to compute total vacancies for digital and non-digital occupations for each region.
  - US location variable: county level; counties are aggregated to the Core-Based Statistical Area level (includes metropolitan and “micropolitan” areas).
  - UK location variable: Indeed’s classification of approximately 80 localities, more aggregate than NUTS-3 but representative of commuting zones and roughly comparable to Travel-to-Work Areas (e.g., Greater London, Greater Manchester).

### Normalized categories (NormCats) for cross-country analysis
- Indeed provides vacancies already aggregated by norm-cats—categories standardized by Indeed based on both industry and occupation.
- These norm-cat series are grouped into two series: digital and non-digital jobs.
- Categories that are unclear in nature are dropped prior to classification. The dropped categories are:
  - “NA”
  - “cat_uncategorized”
  - “overall”
  - “uncategorized”
  - “tangible”
  - “cat_community_and_social_service”
  - “military”
- The classification of the remaining norm-cats is displayed in Annex Table 1.2 (included in the source).
- Sample countries available in the Indeed vacancies data with norm-cats: AUS, AUT, BEL, CAN, CHE, DEU, ESP, FRA, IRL, ITA, JPN, LUX, NLD, NZL, and SWE.

### Key methodological links and notes
- Text-matching algorithms used to map job titles to ISCO-08 codes; further classification into digital/non-digital uses the O*NET digital score and country-specific thresholds described in the main text.
- Aggregation choices aim to produce regionally comparable measures of digital vs non-digital vacancies at monthly frequency for regional analysis and standardized norm-cat series for cross-country regressions.

*Source: 1.5 Vacancies Data (sdnea2023003) — STAFF DISCUSSION NOTES, Digitalization During the Covid-19 Crisis: Implications for Productivity and Labor Markets in Advanced Economies, INTERNATIONAL MONETARY FUND*

### 3.2 Firm-level analysis

### 3.2 Firm-level analysis

### Framework and empirical approach
- Cross-country firm-level data set: Worldscope.
- Estimation approach: regress the average within-firm TFP growth between the pre- and post-pandemic periods on ex ante digitalization measures, firm-level controls, and sector and country/state fixed effects.
- Regressions run separately for 2020 and 2021 and separately for the US and European samples due to different digitalization measures:
  - Europe: ICT measure = 2017–19 average of ICT intensity at the country-sector level (share of workers with Internet access).
  - US: ICT measure = 2017–19 average at the firm level (log of computer equipment and software over workers).
- Firm-level controls X include: firm age, pre-pandemic size (log of total assets), and pre-pandemic profitability (earnings over total assets).
- Fixed effects and standard errors:
  - Advanced Europe: country and sector fixed effects; standard errors clustered at the country-sector level.
  - US: state-sector fixed effects; standard errors clustered at the state-sector level.
- Key parameter of interest: β (effect of pre-pandemic ICT intensity on within-firm change in TFP growth).

### Advanced Europe — main regression specification
- Dependent variable: within-firm difference in TFP growth (annual change in log TFP for t=2020 and t=2021 minus pre-pandemic average annual change in log TFP).
- Pre-pandemic average annual change in log TFP defined as:
  - ∆TFP_i,s,c,pre = 0.5( logTFP_i,s,c,2018 − logTFP_i,s,c,2017 ) + 0.5( logTFP_i,s,c,2019 − logTFP_i,s,c,2018 ).
- ICT measure: ICT_s,c (2017–19 average share of workers with Internet access) at country-sector level.
- Results reported in Annex Table 3.2 (coefficients and fit statistics):

  - Pre-pandemic avg share of workers with internet access:
    - t=2020 coefficient: -0.0293  (standard error reported as (.083) 0.06065 in the table)
    - t=2021 coefficient: 0.2027***  (standard error reported as (.083) 0.06065 in the table)
  - Observations: 1,580 (both columns)
  - R-squared: 0.1985 (t=2020), 0.1133 (t=2021)
  - Sector FE: YES (both)
  - Country FE: YES (both)
  - Root mean square error: 8.8951 (t=2020), 9.3949 (t=2021)
  - Note: Standard errors clustered at the country-sector level. *** p<0.01, ** p<0.05, * p<0.1

- Interpretation of reported coefficients:
  - For 2021 the coefficient on pre-pandemic ICT intensity is positive and statistically significant at the 1 percent level (0.2027***), indicating higher pre-pandemic country-sector ICT intensity is associated with higher within-firm TFP growth in 2021 relative to the pre-pandemic average.
  - For 2020 the coefficient is negative (-0.0293) and not statistically significant at conventional levels in the table.

### United States — main regression specification
- Dependent variable and pre-pandemic average defined analogously to Advanced Europe.
- ICT measure: firm-level lagged ICT = log(computer & software / workers).
- Fixed effects: state-sector; standard errors clustered at the state-sector level.
- Results reported in Annex Table 3.3:

  - Pre-pandemic avg ICT intensity:
    - t=2020 coefficient: 0.00073  (standard error (0.0083))
    - t=2021 coefficient: 0.018***  (standard error (0.017))
  - Observations: 440 (both columns)
  - R-squared: 0.5587 (t=2020), 0.5778 (t=2021)
  - State*Sector FE: YES (both)
  - Root mean square error: 0.08903 (t=2020), 0.08284 (t=2021)
  - Note: Standard errors clustered at the state-sector level. *** p<0.01, ** p<0.05, * p<0.1

- Interpretation:
  - For 2021 the firm-level pre-pandemic ICT intensity coefficient is positive and statistically significant at the 1 percent level (0.018***), indicating firms with higher pre-pandemic computer & software intensity experienced larger within-firm TFP growth in 2021 relative to their pre-pandemic average.
  - For 2020 the coefficient is small and not statistically significant.

### Historical link between total factor productivity and digitalization (US, 2011–19)
- Sample: unbalanced US Worldscope, focusing on 2011–19.
- Firms classified each year and sector into frontier (top 5% of TFP distribution) and non-frontier.
- Regression for non-frontier firm TFP growth:
  - ∆TFP_i,s,t = α ∆TFP_frontier_s,t + β gap_i,s,t−1 + γ ICT_i,s,t−1 + δ′X_i,s,t + α_s + ζ_t + ε_i,s,t
  - gap_i,s,t−1 = lagged distance between frontier and remaining firms in each sector-year cell.
  - ICT_i,t−1 = lagged log(computer & software / workers) at firm level.
  - Expectations: α and β > 0 (technology diffusion and catch-up).
  - Also estimated with labor productivity (LP) growth as the dependent variable.

- Results reported in Annex Table 3.4 (coefficients and fit statistics):

  - Dependent variable columns: TFP growth rate | LP growth rate
  - Frontier firm's TFP growth:
    - TFP growth column: 0.1480***  (standard error (0.033))
    - LP growth column: 0.0815*  (standard error (0.0410))
  - Lagged productivity gap:
    - TFP growth: 0.052***  (standard error (0.004))
    - LP growth: 0.0494***  (standard error (0.0102))
  - Lagged Computer & Software/Workers (log):
    - TFP growth: 0.003***  (standard error (0.001))
    - LP growth: 0.00529***  (standard error (0.00104))
  - Total Assets (log):
    - TFP growth: 0.00171***  (standard error (0.0004))
    - LP growth: 0.00278*  (standard error (0.00130))
  - Age:
    - TFP growth: -0.202*  (standard error (0.0006))
    - LP growth: -0.00685***  (standard error (0.0005))
  - Observations: 6,799 (TFP growth), 7,084 (LP growth)
  - R-squared: 0.08 (TFP growth), 0.0891 (LP growth)
  - Sector FE: YES (both)
  - Year FE: YES (both)
  - Root mean square error: 0.0879 (TFP growth), 0.1662 (LP growth)
  - Note: Standard errors clustered at the state-sector level. *** p<0.01, ** p<0.05, * p<0.1

- Interpretation and substantive findings:
  - Positive and significant coefficient on lagged Computer & Software/Workers (log) in both TFP and LP specifications:
    - TFP growth: 0.003*** — higher lagged firm-level ICT intensity is associated with higher subsequent TFP growth among non-frontier firms.
    - LP growth: 0.00529*** — higher lagged ICT intensity is associated with higher labor productivity growth.
  - Significant coefficients on frontier firm growth (0.1480*** for TFP) and on lagged productivity gap (0.052***), consistent with diffusion/catch-up dynamics from frontier firms to non-frontier firms.
  - Firm size (Total Assets log) positively associated with productivity growth; firm age shows negative association in these specifications.

### Working-from-Home model and calibration (Annex 6 highlights)
- Purpose: gauge potential long-term impact of working from home on labor supply via commuting time savings; household model of joint utility for husband and wife, four household types depending on teleworkability.
- Model features:
  - Households maximize joint utility over consumption and leisure; commuting time T_i enters effective leisure.
  - Interior solution conditions imply:
    - A decline in an agent’s commuting time leads to an increase in that agent’s hours worked and a decline in the spouse’s hours worked.
    - An increase in the preference for leisure leads to a decline in the agent’s hours worked and an increase in the spouse’s hours worked.
- Calibration and data:
  - Data source: American Time Use Survey (ATUS); pre-pandemic averages = 2016–2019; after the pandemic refers to 2021 numbers.
  - Model has 13 parameters: nine chosen exogenously; four calibrated endogenously.
  - Exogenous choices include:
    - Frisch elasticity of labor supply = 3.
    - Average men's and women's commuting time from ATUS (to match pre-pandemic commuting).
    - Wages for each occupation chosen to match wage premium between men and women in teleworkable and non-teleworkable occupations from CPS 2019.
  - Endogenous calibration: husband's and wife's disutility of work (occupation-dependent) calibrated to match average hours worked in each occupation by gender before the pandemic.
- Pandemic-induced shocks considered:
  - Commuting time shock: replicates reduction in commuting time observed across teleworkable occupations (from ATUS, Annex Table 6.1).
  - Leisure shock: replicates the changes in hours worked experienced by men in non-teleworkable occupations after the pandemic; applied to men in both teleworkable and non-teleworkable occupations.

- Annex Table 6.1 — Hours Spent Working and Commuting (selected reported values)
  - Before the Pandemic / After the Pandemic (Working / Commuting / Diff. Working / Diff. Commuting)
  - Women, Tele:
    - Before: Working 35.5, Commuting 2.8
    - After: Working 35.7, Commuting 1.8
    - Diff. Working 0.2, Diff. Commuting -1.0
  - Women, Non-tele:
    - Before: Working 30.6, Commuting 2.4
    - After: Working 30.6, Commuting 2.4
    - Diff. Working 0.1, Diff. Commuting 0.1
  - Men, Tele:
    - Before: Working 43.2, Commuting 4.1
    - After: Working 41.3, Commuting 2.1
    - Diff. Working -1.9, Diff. Commuting -2.1
  - Men, Non-tele:
    - Before: Working 42.4, Commuting 3.9
    - After: Working 43.3, Commuting 3.7
    - Diff. Working 0.9, Diff. Commuting -0.2
  - Sources for table: ATUS; and IMF staff calculations.
  - Note: Working and commuting before the pandemic consist of the average time spent on these activities between 2016 and 2019; after the pandemic refers to 2021 numbers.

### Key empirical implications (from firm-level and historical analyses)
- Pre-pandemic digitalization (ICT intensity) is positively associated with within-firm TFP growth in the post-pandemic period, especially evident in 2021:
  - Advanced Europe: positive and significant effect in 2021 (0.2027***).
  - US: positive and significant effect in 2021 (0.018***).
- Historical (2011–19 US) evidence indicates firm-level ICT (computer & software per worker) positively predicts subsequent TFP and labor productivity growth among non-frontier firms (0.003*** for TFP; 0.00529*** for LP), consistent with ICT fostering productivity improvements and catch-up dynamics.
- Frontier firms’ performance and the lagged productivity gap matter for non-frontier firms’ growth, consistent with diffusion and catch-up mechanisms.

*Source: sdnea2023003 - 3.2 Firm-level analysis (IMF staff discussion notes).*

### Annex 7. Box A1

### Annex 7. Box A1

### Overview of Next Generation EU (NGEU) and the Recovery and Resilience Facility (RRF)
- The NGEU is an EU-wide program providing funding to EU member states subject to the implementation of reforms and investment plans aligned with EU guidance over the period 2021–26.
- The aim of NGEU is to accelerate the recovery of member states and boost their resilience, convergence, and modernization.
- The majority of NGEU funds will be disbursed under the RRF, which accounts for approximately 90 percent of the program.
- This box focuses on proposed RRF grants earmarked for digital transformation investments.

### Size and coverage of proposed RRF digital support
- Analysis uses EU country survey data based on country proposals for RRF digital funds from 15 EU countries.
- Size of proposed RRF digital support ranges from:
  - approximately 1.3 percent of 2020 GDP (Greece)
  - to 0.08 percent of 2020 GDP (Ireland)

### Types of digitalization measures funded
- Public sector-targeted investments proposed under National Recovery and Resilience Plans include:
  - enhancement of e-government and digitalization of administration processes
  - improvement of digital infrastructure for schools
  - increase in cybersecurity and cloud and data processing
  - enhancement of digital infrastructure to improve reliability of the digital business environment
- Private sector-targeted measures include:
  - funding support to firms for development of digital technologies and digital-related R&D (examples given: artificial intelligence, automation, and blockchain)
  - support for the absorption of digital products, services, and processes, with a strong focus on SMEs

### Relationship between prior digitalization and RRF digital allocations
- Countries that were less digitalized prior to the pandemic have the most ambitious digital transformation grant allocation proposals, suggesting efforts for digital catch-up.
- There is a negative relationship between size of RRF digital funds over GDP proposals and the share of workers with internet access prior to the pandemic (reference: Box Figure A1, panel 2), consistent with NGEU’s aim to provide critical support to lagging countries.

### Implementation conditions and broader reform linkages
- Catalyzing NGEU funding is tied to implementation of broad-ranging reforms that aim to upgrade national capacity, covering:
  - labor and product market reforms
  - pension system reforms
  - social inclusion
  - climate policy and energy transition
  - education
  - health care
- Leveraging the large policy support together with needed reforms is identified as key for addressing structural challenges that emerged during the pandemic and for paving the way to convergence and accelerated economic development.

*Prepared by Myrto Oikonomou.*

---


_Source: https://www.imf.org/-/media/files/publications/sdn/2023/english/sdnea2023003.pdf_