Human Capital, Reimagined

The university keeps the transcript, the employer keeps the performance record and the platform keeps the ratings. This series asks whether human capital can work as an open network organised around the person rather than the institutions they pass through. An evidence review neither confirmed the idea nor killed it: nearly every component already exists, the connections between them are the unsolved part, and the first thing worth building is narrow. It is trusted evidence of real work that a second employer can reuse. Nine shifts, each with its evidence, its risks and the test that could end it.

Digital infrastructure has already changed human capital, one layer at a time. Online learning put courses within reach of anyone with a connection. Digital credentials made achievements portable between systems. Skills frameworks and matching tools made candidates searchable, and some public employment services now return tailored job information in seconds rather than minutes^[2]. Each layer improved one piece. None changed the underlying arrangement, in which every institution holds its own piece of a person’s capability.

A person may learn from a university, pick up a micro-skill from a practitioner, practise through an apprenticeship, demonstrate capability on a project, be assessed by an independent evaluator, and then apply that capability in a completely different organisation. Each piece is experienced as a separate system. The person moves across all of them.

The original hypothesis was therefore not to build another learning platform, recruiting marketplace or credential system^[CN]. It was to ask a different question.

What if the infrastructure around human capability were organised around the person and their evolving capability, rather than around the institutions through which that capability happens to be formed?

In the proposed system, the person supplies the intent: I want to become capable of X. Their personal AI agent, not the network, works out the pathway: what to learn and practise, which micro-skills are missing, where each can be acquired and demonstrated, and what should happen next. The identity that results is not another résumé. It is a permissioned, verifiable record that the person controls^[CN]. That changes the direction of the labour-market transaction.

The worker doesn’t disappear from the transaction. They keep control over availability, evidence, rate and whether to accept an opportunity.

Scope. The series concerns knowledge work and the paid work that forms and proves capability. It does not argue that universities, apprenticeship systems, employment services, recruiters or platforms are failing; each solves the part of the problem it was built for. The question is about the connections between them. Most of the evidence comes from formal labour markets in the United States and Europe, so how it transfers to economies with large informal sectors is treated throughout as an open question.

Each participant sees a partial view of the same person.

So the market transacts in stand-ins. An employer with a vacancy writes a job description, people select themselves into applying, and the employer tries to reconstruct capability from degrees, résumés, tests and interviews. Only after hiring does work reveal what the person can actually do. That evidence then mostly stays inside the employer that produced it, and the next employer starts again.

We don’t transact in capability. We transact in representations of capability.

This is not a small inefficiency. Uncertainty about what a candidate can do forces employers into costly screening and leaves capable people unhired, and the cost falls hardest on people with no references or track record^[1]. Matching is slow and expensive, and it suffers from incomplete information and bias^[2].

The evidence review sorted the architecture into what survived and what had to change. Three things survived. Learning has to connect to productive work: the programmes with the strongest results combine training with employer relationships, placement and support^[4]. Apprenticeship can pay for employers, under conditions (Shift 4). And demonstrated work is stronger evidence than a claim about learning (Shift 2).

Four things changed. A capability graph can’t simply be a better résumé, so it should grow only where it improves a transaction. The living identity should hold the evidence a person chooses to use, not everything known about them. Discovery without applications remains a hypothesis. And finance moves from the foundation to an expansion.

The review’s most important finding is the smallest economic object that survives the evidence: trusted, reusable, job-relevant capability evidence. It is not a credential, a graph or a marketplace. It is proof of real work that a second buyer can act on instead of repeating the test. That gives the architecture a sequence.

The research did not kill the idea. It showed where the idea has to become precise.

One capability family with observable output, a small consortium of employers with real paid projects, and an accountable intermediary. Measured in money rather than registrations, and run as two separate bets so a strong result in one cannot hide a weak result in the other.

Established

Nearly every component exists, and several systems join four or five layers in bounded settings; employer uncertainty about ability is a documented source of labour-market inefficiency [1].

Evidence-supported

The interfaces between components, not the components themselves, are the unsolved part.

Hypothesis

Trusted, reusable evidence of real work can lower the cost of later transactions enough to sustain an open coordination layer.

Unknown

Whether employers will reuse and pay for evidence produced elsewhere, and how the architecture transfers to largely informal labour markets.

1. [1]Kässi and Lehdonvirta, Do Microcredentials Help New Workers Enter the Market? Evidence from an Online Labor Platform, Journal of Human Resources 59(4) (2024)Checked
2. [2]Broecke, Artificial Intelligence and Labour Market Matching, OECD Social, Employment and Migration Working Papers (2023)Checked
3. [3]World Bank, Solutions for Youth Employment, draft note on digital job-matching platforms in public agencies (2023)Checked (draft discussion note)
4. [4]Katz, Roth, Hendra and Schaberg, Why Do Sectoral Employment Programs Work? Lessons from WorkAdvance, Journal of Labor Economics (2022)Checked
5. [5]National Governors Association, resource page on learning and employment recordsChecked
6. [CN]Open Commerce for Human Capital: concept document consolidated from the discussion (internal)Provided

Spotted a source we missed or a rule we misread? Tell us what we got wrong.