GCRI · Labs
Future of work lab
Work changes when technologies, contracts, skills and institutions change, and research on that change is often loud and weakly identified. This lab studies how automation, platform work, remote work, skills systems and workplace institutions reshape jobs and organisations, with designs that can tell a real effect from a trend. This page sets out the mechanisms, the measures, the data, the methods and the people who decide on the evidence.
Future of work lab on the record.
Future of work lab: how the risk develops.
Workforce and institutional change research. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Tasks change before jobs
Technologies automate or augment specific tasks, not whole occupations. Jobs are rebundled, skill demands shift within roles, and the effect on employment depends on how firms reorganise work around the tool.
Exposure is not impact
Indices of which occupations are exposed to a technology describe potential, not outcome. Adoption, cost, regulation and bargaining decide what happens, so exposure indices need causal evidence before they inform policy.
Institutions mediate change
Collective agreements, training systems, social protection and labour law shape whether change raises productivity and wages or shifts risk onto workers. The same technology can produce different outcomes under different institutions.
Workplace trials carry power
Experiments inside firms involve employees who may not feel free to refuse. Consent, worker representation, data protection and a real right to opt out decide whether a workplace study is ethical and its results credible.
Stocks, flows, delays and a tipping threshold.
Two stocks, pressure and adaptive capacity, joined by a reinforcing growth loop, a delayed balancing response and damage when pressure outruns capacity; past the threshold the system can tip. Teaching model with stated illustrative values, not a prediction.
Controls
- Growth of pressure
- 6.0% a year
- Efficacy of capacity in reducing pressure
- 4.0% a year
- Investment response to perceived pressure
- 20%
- Perception delay
- 4.0 years
- Tipping threshold of pressure
- 60 index
- Amplification past the threshold
- 2.0 times
- Years simulated
- 50 years
Readouts
- Peak pressure
- 113 index
- Capacity at the end
- 186 index
- Years over the threshold
- 15.8 years
Numbers, methods and sources
| Year | Pressure | Capacity | Perceived |
|---|---|---|---|
| 0 | 20.0 | 50.0 | 20.0 |
| 4 | 23.3 | 44.9 | 21.2 |
| 8 | 27.2 | 42.2 | 24.0 |
| 12 | 31.7 | 42.3 | 27.7 |
| 16 | 36.8 | 45.5 | 32.2 |
| 20 | 42.3 | 51.5 | 37.2 |
| 24 | 47.9 | 60.3 | 42.6 |
| 28 | 53.2 | 71.3 | 48.0 |
| 32 | 57.8 | 84.3 | 53.1 |
| 36 | 66.4 | 98.4 | 58.4 |
| 40 | 82.7 | 115.9 | 69.7 |
| 44 | 97.8 | 139.7 | 83.8 |
| 48 | 109.1 | 169.4 | 97.3 |
Methods
- System dynamics Forrester (1961), Industrial Dynamics, MIT Press; Sterman (2000), Business Dynamics
- Critical transitions and tipping points Scheffer and others (2009), Early-warning signals for critical transitions, Nature 461
- Causal loop diagram Sterman (2000), Business Dynamics, McGraw-Hill
- Uncertainty fan (parameter Monte Carlo) Spiegelhalter, Pearson and Short (2011), Visualising uncertainty about the future, Science 333
- One-at-a-time sensitivity (tornado) Saltelli and others (2008), Global Sensitivity Analysis: The Primer, Wiley
Data
No observed data enters this model: it runs on the teaching values set by the controls, to be replaced with a programme’s own evidence.
Across the six pillars, led by Nexus Consortium
- Experiments Labs in Labs
- Built in Labs in Foundry
- Tested at Labs in Universe
- Codified in Data and Evidence in Standards
- Published in Labs in Reports
- Taught in Faculty of Risk Science and Methods in Academy
The work lab studies change in each domain’s workforce.
Findings feed the Labs Evidence series. The ILO’s 2023 working paper sets out exposure measures.
- Workplace study sites.
- Worker accounts with consent.
How future of work lab is modelled, measured and tested.
The model, its assumptions, its data and its uncertainty, in the form a scientist or engineer would check. Structures follow published methods. The values drawn are illustrative, not findings.
Model, equations and assumptions
The model structure used for future of work lab, the assumptions it makes and how each is tested.
Ri = f(Hi, Ei, Vi, Ci)hazard, exposure, vulnerability and coping capacityEAL = ∑k pk · Lkover events k with annual probability pkxt+1 = A xt + B ut + εtA carries the couplings, u the shock, ε the error| Assumption | Why it is made | How it is tested |
|---|---|---|
| Hazard and exposure are independent within an event | Keeps the loss integral tractable | Compare with a copula fit on joint records |
| Couplings are linear near the operating point | Lets shocks propagate through A | Perturb beyond the linear range and compare |
| Vulnerability curves transfer between places | Local curves are rarely available | Validate against observed losses where they exist |
- Inputs
- versioned datasets with provenance
- Outputs
- distributions, never single numbers
- Code
- released with the model card and tests
- Reproduction
- same data, code, environment and seed
What is measured, and how well.
Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Task exposure index | dimensionless index by occupation | annual | Task databases mapped to technology capability | MODELLED |
| Firm adoption of the studied technology | percent of firms | annual | Enterprise surveys | ESTIMATED |
| Adult skills proficiency | score on the survey scale | per survey cycle | OECD Survey of Adult Skills | OBSERVED |
| Job quality index | composite index | per survey cycle | Working conditions surveys | ESTIMATED |
| Participation in job-related training | percent of adults | annual | Labour force and adult education surveys | Published statistics |
| Pre-analysis plans for workplace trials | percent of trials registered before launch | annual | AEA RCT Registry | Published statistics |
| Worker representatives consulted | percent of workplace studies | per study | Study ethics records | DECLARED |
Where the evidence comes from.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| ILOSTAT (International Labour Organization) | Employment, hours, wages and informality by country | Free and open, CC BY |
| OECD Survey of Adult Skills (PIAAC) | Literacy, numeracy and problem solving of adults, with job tasks | OECD terms; microdata by release |
| O*NET (US Department of Labor) | Task, skill and work activity descriptors by occupation | Open, CC BY |
| ESCO (European Commission) | Multilingual classification of skills, competences and occupations | Free and open, attribution |
| European Working Conditions Survey (Eurofound) | Job quality, autonomy, work intensity and representation | Free for research, end-user licence |
| AEA RCT Registry (American Economic Association) | Registered field experiments and pre-analysis plans in economics | Free access, terms of use |
| National labour force surveys | Employment, occupation and training microdata | Terms per statistical office |
Who decides, and what the evidence serves.
Labs prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Labour ministries and public employment services | Which training and transition programmes to fund or scale | Impact evaluations with credible comparison groups, cost data, effects by group |
| Employers and worker representatives | Whether and how to introduce a technology, with what consultation and safeguards | Pilot evaluation, task and job-quality effects, data protection assessment |
| Research ethics committees | Whether a workplace experiment protects voluntariness and personal data | Consent design, opt-out route, representation agreement, data minimisation |
| Research funders | Which questions on work and technology merit sustained funding | Evidence gaps mapped by strength of design, not by volume of commentary |
Outputs this topic produces
- Pre-analysis plans for workplace and labour market trials
- Occupational task maps linked to technology capability, with the method stated
- Evaluations of training and transition programmes with the identification strategy stated
- Ethics and consultation records for studies inside workplaces
What experts ask about changing work.
- Which tasks within an occupation are changing, rather than which jobs disappear?
- Who bears the costs of transition, by region, age, sex, sector and skill level?
- Which training, matching and income-support measures actually change outcomes?
- How do institutions such as schools, unions, employers and public employment services adapt?
- How do heat, disasters and supply disruptions affect work and safety?
The forces reshaping work.
Economists analyse technological change at the level of tasks. A technology may substitute for some tasks, complement others and create new ones; the net effect on an occupation depends on the mix and on how employers reorganise work. Automation and AI tend to affect routine cognitive and manual tasks first, but recent tools also reach into tasks once thought to require judgement.
Other forces act alongside technology. Ageing populations change labour supply and care needs. Climate change raises heat exposure for outdoor and industrial workers and shifts employment between sectors during energy transitions. Trade and supply-chain disruptions can concentrate job losses in particular places, where the effects persist for years.
Institutions shape outcomes. The same technology can produce very different results depending on training systems, labour law, social protection and bargaining arrangements.
Evidence and its limits.
- 01
Task and occupation data
Classifications of work activities linked to employment data. They lag behind real change and vary between countries.
- 02
Labour force and administrative records
Surveys and registers tracking employment, earnings and transitions. Access is often restricted and linking sources raises privacy questions.
- 03
Impact evaluation
Randomised and quasi-experimental studies of training and matching programmes using causal inference. Results may not transfer across labour markets.
- 04
Workplace experiments
Controlled trials of new tools or practices inside organisations, under consent and ethics review.
Forecasts of jobs lost or created are highly sensitive to assumptions and are not presented as predictions. The lab reports scenarios with their assumptions.
How it is studied, and what the methods cannot say.
- 01Causal inference and impact evaluationRandomised pilots and stepped rollouts · Quasi-experimental programme evaluationApplied here
- 02Probability and statisticsSample size and power planning · Meta-analysis with bias assessmentApplied here
- 03Artificial intelligence and machine learningAssisted screening for systematic reviews · Surrogate models to guide experimentsApplied here
- 04Uncertainty quantification and model validationMeasurement uncertainty budgets · Sensitivity analysis of simulation modelsApplied here
Where this page’s references lead.
The bodies and documents this page is read against. Status and editions change; each link goes to the issuing body.
Questions and answers.
Does exposure to automation mean jobs will be lost?
Why register a pre-analysis plan for a field experiment?
Can employees refuse to take part in a workplace study?
Does the lab predict how many jobs will be lost to AI?
Can an employer run a workplace trial with the lab?
How does this lab differ from the Future of Work and Skills platform?
Contribute evidence from this page.
Reports is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.
- EmailGateway GW-GCRISend it from this pageEvidence, a correction, a case, a method, a review, a translation or profile input. It arrives tagged with this page, so the editor who owns it receives it.
- ZenodoOpen depositDeposit it and submit to the communityPublish your own record on Zenodo with your ORCID iD, then submit it to the Reports community there. Curators review it; the DOI stays yours.
- ReportsGCRISee every way to contributeThe contribution types, how each is reviewed, what credit you receive, the licence and what is never accepted.
Never send personal data, sensitive infrastructure detail or anything you have no right to share. A contribution is reviewed before use; sending it confers no authorship or editorial rights, and sponsorship never buys a finding.
Bring a method, a dataset or a question.
Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.
Legal documents.
Maps, data and figures on this site are shown for context and general information. They are not advice, forecasts or warnings; the limits of the information are set out in the Disclaimers and Data Notices.