On Work, Usefulness, and What We Measure
Work has a strange way of becoming indistinguishable from worth.
Not because they are the same thing, but because most societies have become very good at measuring one through the language of the other.
We ask what someone does before we ask what they know. We trust titles because they make competence easier to categorise. We treat employment as evidence of usefulness, even though useful work has always existed far beyond payrolls, offices and formal institutions.
A farmer may understand risk better than a consultant. A mechanic may solve problems with more precision than someone whose profession is built around the phrase “problem-solving”. A person without a job may possess exactly the skills an organisation claims to need and still remain invisible to it.
This is not necessarily hypocrisy. It may simply be a limitation of systems.
Large systems need categories. Labour markets need qualifications, titles, industries and experience thresholds because millions of people cannot be evaluated as individuals every time a decision is made. The problem begins when those categories stop being tools for understanding people and become substitutes for understanding them.
That tension is particularly visible in unemployment.
An unemployment rate describes people without employment who are seeking and available for work, under a stated statistical definition. It is useful, but it cannot describe the full range of work and knowledge outside paid employment.
It tells us very little about the skills sitting idle, the work being performed outside formal employment, the mismatch between what institutions say they need and what they are willing to recognise, or the strange process through which a capable person can be classified as outside employment without becoming any less capable.
So I decided to look at the numbers more closely.
Beyond Employment examines what being employed tells us about economic security in Kenya. It draws on household survey data from 2015/16 and public-opinion data from 2024, looking at household poverty, working conditions and experiences of material hardship.
The data cannot tell us how much useful knowledge goes unrecognised, or how often institutions overlook capable people. It can help examine a narrower question: how much we can reasonably infer about someone’s circumstances from knowing that they have work.
Because perhaps the more interesting question is not whether people are willing to work.
It is how well our systems recognise where useful work, useful knowledge and useful people already exist.