Cohort profile: Description of the GIG-OSH longitudinal cohort on occupational safety and health of digital platforms workers in Europe

The GIG-OSH cohort is the first large-scale, longitudinal study across seven European countries designed to investigate the occupational safety, health, and working conditions of digital platform workers, revealing significant mental health disparities and high work intensities while highlighting methodological challenges related to non-probabilistic sampling and data harmonization.

The Gist

Imagine the modern workforce as a massive, bustling marketplace. For decades, this marketplace had clear stalls: you worked in an office, a factory, or a shop, and you had a boss, a fixed schedule, and a paycheck.

But recently, a new, digital marketplace has exploded onto the scene. This is the "Gig Economy," run by apps like Uber, Deliveroo, Upwork, and Fiverr. Instead of a boss, you have an algorithm (a computer code) that assigns you tasks. You might be a food delivery rider racing through traffic, or a graphic designer working from a coffee shop halfway across the world.

This paper is the "report card" for a massive new study called GIG-OSH. Its goal? To figure out how working in this digital marketplace affects people's safety, health, and happiness across seven European countries.

Here is the story of the study, broken down simply:

1. The Mission: A Long-Term Health Check

Think of the researchers as a team of doctors trying to understand a new type of patient. They didn't just want to ask, "Do you have a headache?" once. They wanted to follow these workers over time to see if the "digital job" causes long-term wear and tear.

They gathered 3,945 workers from Belgium, Denmark, Finland, Poland, Spain, Sweden, and the UK. They split them into two main groups:

• The "On-Location" Crew: The runners, drivers, and handymen who have to be physically present (like a pizza delivery rider).
• The "Web-Based" Crew: The freelancers, translators, and data entry clerks who work entirely online.

2. The Recruitment: Finding the Invisible

Finding these workers was like trying to find ghosts in a crowd. There is no official list of gig workers. You can't just look up "Uber drivers" in a phone book because many are independent contractors, not employees.

So, the researchers used a "fishing" strategy with many different nets:

• They put up digital billboards on social media (Facebook, Instagram).
• They hung out in online communities where workers chat.
• They went out into the streets to talk to riders at traffic lights.
• They even used snowball sampling, asking one worker to tell their friends to join.

3. What They Found: The Good, The Bad, and The Ugly

Once they got the data, they looked at the workers' lives. Here is what the "report card" showed:

• The "Two Worlds" Divide:

  • The Web-Based workers are mostly educated, often have other jobs, and work fewer hours. They have more freedom (autonomy) but often feel isolated and earn very little money from the platform.
  • The On-Location workers are more likely to be men and immigrants. They work a lot of hours (often 78 hours a month just on the app) and rely heavily on this income. They face physical dangers like traffic accidents and bad weather.

• The Mental Health Gap:
  Imagine a "Well-Being Battery." The average European person has a battery charge of 69.4%. The gig workers in this study only had 58.7%. They are running on lower battery power. They are more stressed, likely because they are constantly worried about their next task, their rating, or their income.

• The "Algorithm" Boss:
  Unlike a human boss who might say, "Take a break," the algorithm is a relentless machine. It tracks every move, assigns tasks instantly, and can "fire" you (deactivate your account) if your rating drops. This creates a unique kind of stress called algorithmic management.

• The Safety Gear Problem:
  If you are a delivery rider, you need a helmet. The study found that while some platforms provide gear, many workers have to buy their own or go without. For web-based workers, the "safety gear" is non-existent because their risks are digital (like cyberbullying or data theft), which people often forget to protect against.

4. The Challenges: Why This Was Hard

The researchers admitted the study wasn't perfect.

• The "Dropout" Rate: Because gig work is unstable, many workers dropped out of the study after the first survey. It's like trying to study a group of people who are constantly moving houses; it's hard to keep track of them.
• The "Snapshot" Issue: Since they couldn't get a perfect list of all workers, they couldn't say, "This represents 100% of all gig workers." It's more like a very detailed photo of a specific crowd, rather than a census of the whole city.

5. The Big Picture: Why Does This Matter?

This study is a wake-up call. For a long time, people thought "flexible work" was a dream come true. This study suggests that for many, it's a double-edged sword. It offers freedom, but it often comes with high stress, low pay, and a lack of safety nets.

The Takeaway:
The GIG-OSH study is like a map for the future. It shows us that as more of us move to this digital marketplace, we need new rules. We need to figure out how to protect the "ghosts in the crowd"—ensuring they have fair pay, mental health support, and safety gear, whether they are delivering food or designing logos.

The researchers promise to keep watching, hoping to build better tools to measure exactly how much these workers are really earning and how much time they spend just "waiting" for the next job. Until then, this study is the most important compass we have for navigating the future of work.

Technical Summary

1. Problem Statement and Research Gap

Digital platform work (DPW) has expanded rapidly in Europe, characterized by algorithmic management and task-based remuneration. While existing literature highlights the precarious nature of this work (low wages, lack of benefits, algorithmic control), there is a critical evidence gap regarding its long-term impact on Occupational Safety and Health (OSH).

• Limitations of prior research: Most studies are qualitative, country-specific, or cross-sectional. There is a scarcity of large-scale, longitudinal data comparing different types of platform work (web-based vs. on-location) across multiple national contexts.
• Objective: To establish the GIG-OSH cohort, the first large-scale longitudinal study in Europe designed to investigate the OSH, working conditions, and health outcomes of digital platform workers across seven countries with distinct welfare regimes.

2. Methodology

Study Design and Population

• Design: Prospective longitudinal cohort study with a baseline survey and follow-up waves (6 months for all countries; 12 months for Spain).
• Countries: Belgium, Denmark, Finland, Poland, Spain, Sweden, and the United Kingdom.
• Target Population: Digital platform workers earning income via apps/websites, categorized into:
  • On-location: Physical tasks (e.g., delivery, transport, household services).
  • Web-based: Remote tasks (e.g., micro-tasks, freelancing, data entry).
• Sample Size:
  • Baseline: 3,945 participants (after exclusions for non-eligibility and data cleaning).
  • Follow-up (Wave 2): 385 participants (significant attrition noted).

Recruitment Strategies

Due to the lack of a central registry for platform workers, the study employed non-probabilistic, purposive sampling tailored to national contexts. Strategies included:

• Targeted social media advertising (Facebook, LinkedIn, etc.).
• Organic outreach via worker networks (Telegram, Facebook groups).
• Recruitment via microtask platforms (Upwork, Prolific).
• Fieldwork/in-person recruitment at transport hubs.
• Institutional collaboration (unions, universities) and snowball sampling.

Data Collection and Instruments

• Platform: Online surveys administered via REDCap.
• Variables:
  • Demographics: Age, gender, migration status, education.
  • Work Conditions: Task type, contract status, working hours, income (relative to national median), algorithmic management exposure.
  • Health Indicators: Self-perceived health, WHO-5 Well-Being Index, physical symptoms, and specific OSH risks (traffic, psychosocial, cyberbullying).
  • Composite Indices: Derived from the European Working Conditions Survey (EWCS) methodology, including indices for earnings, contract type, working time quality, work intensity, physical environment, social environment, and autonomy.
• Statistical Analysis: Descriptive statistics and Non-linear Canonical Correlation Analysis (HOMALS) to map relationships between task types and employment conditions.

3. Key Results (Findings to Date)

Demographics and Labor Characteristics

• Composition: The cohort is predominantly male (58.8%) with a mean age of 32.6 years.
• Task Segmentation:
  • On-location workers: More likely to be male (69.2%), migrants (41.2%), and engaged in delivery services (70.8%). They work significantly more hours (mean ~78 hours/month) and earn a higher percentage of the national median income (29.5%) compared to web-based workers.
  • Web-based workers: More likely to be female, have tertiary education (64.4%), and engage in micro-tasks (52.7%). They work fewer hours (mean ~29 hours/month) and earn a lower percentage of the national median (12.6%).
• Precarity: A majority of web-based workers (56.2%) lack a formal employment contract, compared to 25.7% of on-location workers.

Occupational Safety and Health (OSH)

• Mental Health: The mean WHO-5 Well-Being Index score was 58.7, significantly lower than the European general population average (69.4), indicating poorer mental health outcomes among platform workers.
• Physical Risks: On-location workers face higher physical demands and traffic risks. Only ~20% of on-location workers reported receiving protective equipment paid by the platform; 74% of web-based workers reported using no protective equipment.
• Working Conditions:
  • On-location: Higher work intensity, poorer working time quality, and more demanding physical environments.
  • Web-based: Higher autonomy and better social environments, but lower earnings and higher rates of "no contract" status.

Multidimensional Analysis (HOMALS)

The analysis revealed two primary dimensions structuring the data:

1. Task Type & Income: A clear separation between remote micro-tasks (low income, low hours, low dependency on platform income) and on-location delivery/transport (higher income, higher hours, higher dependency).
2. Employment Arrangement: A distinction between self-employment (common across both types) and formal/informal contracts. Formal contracts were more associated with high-commitment on-location tasks, while "no contract" status was prevalent among micro-taskers.

• National Variation: The UK, Belgium, and Spain were heavily associated with remote micro-tasks, whereas Denmark and Poland were dominated by delivery workers.

4. Key Contributions

• First Large-Scale Longitudinal Cohort: Establishes the first multi-country, longitudinal dataset specifically tracking OSH in the gig economy, addressing a major gap in occupational health literature.
• Comparative Scope: Enables direct comparison between web-based and on-location workers, revealing distinct risk profiles and employment conditions that single-profile studies miss.
• Stakeholder-Informed Design: The questionnaire was co-created with workers, unions, and policymakers, ensuring relevance to real-world issues like algorithmic management and unpaid waiting time.
• Methodological Innovation: Successfully recruited a "hard-to-reach" population using mixed recruitment strategies across diverse national welfare regimes.

5. Limitations

• Sampling Bias: Non-probabilistic sampling limits the statistical representativeness of the cohort for national populations; weighting against labor force surveys was not possible due to a lack of baseline data on platform workers.
• Attrition: High dropout rates between waves (from 3,945 to 385) constrain the power of longitudinal analyses and generalizability of long-term trends.
• Self-Reported Data: Reliance on self-reported measures introduces potential recall and social desirability biases.
• Measurement Gaps: Current tools struggle to accurately capture hourly wages and unpaid waiting time, which are critical for assessing economic precarity.
• Harmonization: Differences in national classification systems (e.g., education levels) required adaptations that may affect cross-country comparability.

6. Significance and Future Directions

The GIG-OSH cohort provides robust empirical evidence that platform work is not a monolith; it comprises distinct subgroups with divergent health risks and employment conditions.

• Policy Relevance: The findings support the need for differentiated regulatory approaches for web-based vs. on-location workers, particularly regarding social protection, algorithmic transparency, and safety standards.
• Future Plans:
  • Develop precise metrics for hourly earnings and unpaid waiting time.
  • Expand recruitment to underrepresented groups (e.g., medical and domestic care platform workers).
  • Link cohort data with administrative records to track long-term health trajectories and the impact of new EU labor regulations.

This study serves as a foundational resource for researchers, policymakers, and stakeholders aiming to understand and mitigate the occupational risks of the digital gig economy in Europe.