0 cumulative citations
View corpus contextAdvertised demand for data skills clusters in the same occupations across the US, UK and Canada, but its industrial footprint varies — suggesting common skill complements to data assets alongside country-specific labor-demand structures.
Mapping Labor Demand for Data‐Intensive Work in the United Kingdom, Canada, and the United States
openalex
descriptive
medium evidence
7/10 relevance
Summary only
summary available; pdf_status=paywall
DOI
Source
PDF
The paper develops an NLP method to identify data-intensive skills in job ads and uses it to show that while occupations demanding data expertise are similarly ranked across the US, UK and Canada, the industry-level distribution of those jobs differs by country.
Citation observations
Cumulative provider counts captured on specific dates; providers are never combined.
0 cumulative citations
View corpus contextABSTRACT The paper makes three main contributions. First, it develops an NLP‐based methodology to identify data‐intensive skills in job advertisements at scale. Second, it operationalizes a transparent indicator of data intensity for occupations and industries. Third, by applying the method to job advertisements from the United Kingdom, Canada, and the United States, it provides harmonized cross‐country estimates of data‐expert hiring, offering a key input to derive a proxy for investment in data assets. Empirical results show that although the ranking of data‐intensive occupations is broadly similar across countries, the industrial distribution of data‐intensive jobs differs, reflecting distinctive labor‐demand structures.
Summary
Main Finding
The paper introduces a reproducible, NLP-based approach to identify data-intensive skills in online job advertisements and uses it to build a transparent indicator of data intensity by occupation and industry. Applied to job ads from the United Kingdom, Canada, and the United States, the measure yields harmonized cross‑country estimates of hiring for data experts. Results show similar rankings of data‑intensive occupations across countries but meaningful differences in the industrial distribution of those jobs, reflecting distinct national labor‑demand structures.
Key Points
- Three core contributions:
- An NLP methodology to identify data‑intensive skills in job advertisements at scale.
- A transparent, operational indicator of data intensity for occupations and industries.
- Cross‑country application (UK, Canada, US) producing harmonized estimates of data‑expert hiring, usable as a proxy for investment in data assets.
- Cross‑country comparison findings:
- Occupational rankings of data intensity are broadly consistent across the three countries.
- The industry-level composition of data‑intensive jobs varies across countries, driven by differing labor‑demand structures and sectoral specialization.
- The indicator is intended to be a practical input for measuring data‑related investment and monitoring labor market changes associated with the data economy and AI adoption.
Data & Methods
- Data: Large corpora of online job advertisements from three countries (United Kingdom, Canada, United States). (The abstract does not specify exact sources or time coverage.)
- Methodology:
- Natural Language Processing (NLP) is used to detect mentions of data‑intensive skills and roles within job ad text.
- Identified skills are aggregated to construct a transparent, reproducible indicator of “data intensity” at the occupation and industry level.
- Cross‑country harmonization procedures map job ads to comparable occupational/industry categories to enable international comparison.
- Outputs:
- Occupation‑level and industry‑level measures of hiring demand for data experts.
- Harmonized metrics that can be used as a proxy for investment in data assets and inputs into broader measures of digital/AI capital.
Implications for AI Economics
- Measurement:
- Provides a scalable, timely proxy for investment in data assets—an important complement to traditional capital‑formation statistics that typically miss intangible data investments.
- Enables cross‑country and sectoral monitoring of labor demand for data expertise, improving empirical work on AI diffusion and data capital.
- Policy and labor markets:
- Helps identify which occupations and industries will drive demand for data skills, informing reskilling and education policy.
- Reveals sectoral differences that matter for targeted workforce development and for anticipating distributional impacts of AI.
- Research and evaluation:
- Can be combined with wage, productivity, firm‑level, and adoption data to study returns to data capital, complementarity between data and AI, and substitution/complementarity between labor and AI technologies.
- Offers a basis for longitudinal studies of how data‑expert hiring evolves with AI adoption and regulatory changes.
- Limitations to consider:
- Reliance on job ad data may undercount unadvertised hiring or internal training/upskilling.
- Harmonization decisions and NLP classification choices affect comparability; transparency and validation are crucial for credible inference.
Assessment
Paper Typedescriptive
Evidence Strengthmedium — The paper provides large-scale, harmonized descriptive evidence on advertised demand for data-intensive skills using an NLP classifier applied to job ads across three countries; this yields credible bounds on advertised skill demand but does not measure actual hiring, hours, productivity or causal effects, and depends on classifier accuracy and job-board coverage.
Methods Rigormedium — The approach appears methodical—developing an NLP method, operationalizing a transparent indicator, and harmonizing cross-country data—but the abstract does not report validation statistics, robustness checks, temporal coverage, or adjustments for platform and occupational coding differences, which leaves measurement error and comparability concerns.
SampleOnline job advertisements from the United Kingdom, Canada, and the United States, processed with an NLP classifier to identify data-intensive skills and aggregated to occupation and industry levels; exact time frame, job-board sources, sample size, and sampling/weighting procedures are not specified in the abstract.
Themesskills_training labor_markets adoption
GeneralizabilityJob advertisements reflect advertised demand, not actual hires or on-the-job tasks, Coverage depends on which job boards/platforms were scraped and may miss employer types that rarely post online (e.g., small firms, internal hires), Cross-country comparability may be affected by differences in posting practices, occupational coding, and language/terminology, NLP classifier accuracy and training data may induce measurement error, particularly for niche or evolving skills, No information on time period means results may not generalize to other years or rapid technological change
Claims (6)
| Claim | Direction | Outcome | Confidence & Evidence | Details |
|---|---|---|---|---|
| We develop an NLP-based methodology to identify data-intensive skills in job advertisements at scale. Other | positive | identification of data-intensive skills in job advertisements |
Reading fidelity
high
Study strength
medium
|
not reported
|
| We operationalize a transparent indicator of data intensity for occupations and industries. Adoption Rate | positive | data intensity indicator for occupations and industries |
Reading fidelity
high
Study strength
medium
|
not reported
|
| By applying the method to job advertisements from the United Kingdom, Canada, and the United States, we provide harmonized cross-country estimates of data-expert hiring, offering a key input to derive a proxy for investment in data assets. Hiring | positive | data-expert hiring (cross-country estimates) |
Reading fidelity
high
Study strength
medium
|
not reported
|
| Empirical results show that although the ranking of data-intensive occupations is broadly similar across countries. Hiring | null_result | ranking similarity of data-intensive occupations across countries |
Reading fidelity
high
Study strength
medium
|
not reported
|
| The industrial distribution of data-intensive jobs differs [across countries], reflecting distinctive labor-demand structures. Market Structure | positive | industrial distribution of data-intensive jobs |
Reading fidelity
high
Study strength
medium
|
not reported
|
| Data-expert hiring (as measured from job advertisements) can serve as a proxy for investment in data assets. Adoption Rate | positive | use of hiring estimates as proxy for investment in data assets |
Reading fidelity
medium
Study strength
speculative
|
not reported
|