In 2018 we opened with a famous forecast that freelancers would be 40% of the workforce by 2020; the number came close only on the broadest definition, but the post's real idea, coworking as entrepreneurship infrastructure for smaller cities, still stands.
In March 2018 we published The Way We Work: Past, Present & Future, written with Anita Fuzi and based on her 2016 dissertation about coworking and entrepreneurship in economically challenged areas.
We opened it with a number: "by 2020, 40% of the work force will be made up of freelancers and solopreneurs." It was the kind of statistic everybody had heard. Looking back, it was also the weakest part of the piece. The strongest part was buried further down. So this time we will deal with the number quickly, and then spend most of our time on the idea that deserved the headline.
Where the 40% came from
The figure goes back to a 2010 forecast by Intuit, which predicted that about 40% of the US workforce would be contingent workers by 2020 (Quartz). In 2017 Intuit's CEO reportedly restated it as 34% at the time and 43% by 2020 (CNN Money).
Two things stand out now. It was a forecast about the US, and we quoted it as if it described the whole workforce everywhere. And "contingent" is a broad word that can mean very different things depending on who is counting.
What the data shows now
Whether the 40% "came true" depends entirely on the definition you use.
- On the broadest definition, roughly. Upwork's own surveys, which use a broad definition of freelancing, put the share at 36% in 2020 (a figure we could not re-check) and 38% in 2023 (Upwork).
- On the official definition, not even close. The US Bureau of Labor Statistics counted independent contractors at 7.4% of workers in 2023 (BLS).
So the honest answer is that 40% was never real on any official measure. It only came near on the broadest survey definition.
Our UK number has not aged well either. We wrote that there were 4.6 million self-employed people in the UK, "and this population is continuing to grow". It did not keep growing. UK self-employment fell sharply from 2020. ONS figures show about 4.44 million self-employed people in May–July 2026, down from a peak of about 5.02 million in late 2019 (ONS). That is lower than the figure we quoted in 2018.
And our closing line, that "the potential for coworking spaces is limitless", reads differently after WeWork, the best-known name in the sector, went bankrupt in 2023 (WeWork). Potential is never limitless. Business models still have to work.
The idea that deserved the headline
Strip away the statistics and the 2018 post was making a much more interesting argument, drawn from Anita's research. Here it is in plain terms.
Most coworking research looked at big cities
Studies of coworking and entrepreneurship mostly looked at places like London, New York and Amsterdam. Smaller cities were left out, even though they are where new businesses often struggle most: fewer funding opportunities and fewer professional resources.
A coworking space can be infrastructure
The argument was that a well-run shared workspace is more than desks. In a smaller city it can act as entrepreneurship infrastructure: a place where people meet peers, learn skills, find support and get closer to funding they might otherwise not reach.
The soft side matters most
This was the most useful finding. Design gives a space its base value, but the social and cultural side decides whether it works. Earlier efforts to boost local economies, the post argued, had failed by neglecting exactly these softer pieces.
One model does not fit every city
The post's case study was the Cardiff city region in Wales. There, the culture was already supportive and social. What made local spaces succeed was confidence building and skill building. The spaces worked more like learning spaces, and gentle facilitation, which encouraged members to share with each other, did better than more direct, organised facilitation. What works in Paris or Boston might flop in Cardiff or Helsinki, and the other way round.
What we can and cannot say now
We also wrote that "medium-sized cities will soon become the new economic drivers". We have not been able to check that claim against current data, so we will not pretend to give it a verdict. It stays an open question.
What we would still stand behind is the practical advice for anyone running, or starting, a shared space outside a big city:
- Start with your members, not your floor plan. Ask what they lack locally: skills, confidence, contacts, funding.
- Build learning into the space. Peer sessions, show-and-tell, people teaching each other. In Cardiff, that is what made spaces work.
- Match your facilitation to the local culture. Some communities need a gentle nudge, others need more structure.
- Work with local partners. The original post ended by suggesting spaces work with policy makers. That still makes sense, as long as nobody promises "limitless" results.
If you want the background on how coworking spaces differ from incubators and accelerators, our older piece on what's what is still in the archive.
The honest scorecard
- 40% of the workforce freelance by 2020: came close only on the broadest survey definition.
- UK self-employment "continuing to grow": didn't happen.
- Medium-sized cities as the new economic drivers: not verified either way.
- Coworking's potential is "limitless": out of date.
- Community and learning make coworking work, especially in smaller cities: still the part worth reading.
Written 3 November 2026 from the research behind the archive's 2026 updates (checked 30 September, re-verified 3 November 2026). Sources are linked in the text.