The promises of the gig economy “evoke the artist’s life in which each concert, or ‘gig’, is but a one-off task or transaction, without further commitments on either side,” writes Jeremias Prassl in his recent book. For some, the gig economy is the new reality of work and opportunity in our ‘technosociety’. For others, it is a return to power relations that labour law hoped to, and perhaps still can, stamp out. We asked the Oxford legal academic for his take on new forms of work.

Krisztian Simon: What is your definition of the gig economy?

Jeremias Prassl: The gig economy entails every form of work that is mediated through online applications. It includes everything from companies that offer a particular service, such as ‘ride-sharing’ companies, to platforms that span a wide range of services, such as pizza delivery or transportation services. There are many different definitions of the gig economy but they all have two things in common. The first is that platforms use software to connect workers out there in the crowd with consumers who want their services. And the second is that they use clever algorithms to curate and deliver that service – things like rating mechanisms or automated payment systems.

The one-off gigs of these workers resemble, to some extent, the situation of a construction worker who is hired by a company without any long-term commitment. Why would this person not fall into the category of gig worker?

You hit on an important point here, because I don’t think that the gig economy is a special phenomenon in the labour market. The challenges we see in the gig economy just stand as a small example, pars pro toto, of the larger challenges in the labour market. The move away from an open-ended relationship with a single employer has been going on for many years. Nevertheless, I wouldn’t call your example a gig worker, because here the construction worker does not rely on a modern, app-based technology – even if the challenges we see are often the same.

When did the trend away from long-term employment become visible?

Sociologists, economists, and lawyers have been studying the broader economic trend for three or four decades. The gig economy of course, with apps and platforms altering the picture, is fairly new. I don’t think it has been around for more than a decade, if that. Uber or Deliveroo are prime examples of gig economy companies.

And this form of employment has both pros and cons…

At the moment, you find literature which either emphasises the negative sides of the gig economy or is very positive about it. What I am trying to say in Humans as a Service is that there is truth in both these narratives. There are some very negative aspects of some forms of the gig economy, but it also offers some real opportunities for workers and consumers.

What is the extent of gig work these days?

The studies vary, but realistically we are looking at no more than 1 to 3 per cent of the labour market. However, gig work is a global phenomenon. Platforms might have different names in different countries, but they are present in most parts of the world. So even if this 1 to 3 per cent might seem small, it matters much more broadly and it is therefore important for us to think a lot harder about the issue.

Which sectors and professions are most affected by this kind of work?

It is changing. Early on, the gig economy was only based on traditional, service-based sectors – things like cleaning, driving, or transportation. But these days lots of jobs are becoming ‘gigified.’ Lots of apps are even appearing in sectors such as healthcare. So even though it started out in a particular segment of the labour market, it has the potential to affect all of us.

What would be an example of gig work where the workers are really profiting from it?

One story told very widely is that of Uber drivers in France who work in the banlieues, the poor outskirts of Paris. These people have traditionally been excluded from the labour market for different reasons, and this way they are finding a way back in.

But Uber also distorts the economy: it pushes cab drivers out of business and there have also been reports that the platform evades tax.

It is very important that with the gig economy, just as with any other business, we have a level playing field. We have to make sure that everyone who operates in the market complies with the same basic rules and regulations.

With such a global, web-based phenomenon, does it also mean that you must have the same regulation in every country, at least inside the EU?

Not necessarily. One of the paradoxes of the gig economy is that even though it is very international in terms of its business model, the vast majority of its services are delivered quite locally. What is offered locally can also be regulated locally, as the Court of Justice of the European Union decided just before Christmas in regard to the specific case of Uber.[1]

What other positive effects does the gig economy bring?

One thing I already highlighted was getting people into the labour market. Another thing that people in the gig economy mention is flexibility; they enjoy choosing when to work, what kind of job to do, and so on. Even though you have to fit it around your timetable, it provides you with the opportunity to earn extra cash.

But then again, if you have a family, if you have to take your children to school in the morning and be there for them in the afternoon, all these aspects can become very inconvenient.

Flexibility, if it is not managed properly, can also become insecurity. If you try to work at a particular time and nobody is available, then the flexibility of your work becomes illusory.

What other negative aspects are there for workers?

What we see is fairly tight control, including quite significant restrictions, which greatly reduces the scope for entrepreneurship. The rating mechanism is also very problematic for many workers. Supposedly it is only about quality and feedback, but in reality, the algorithm can easily become your new boss, controlling you very tightly, ensuring that you work quickly. Many workers end up working under a lot of pressure.

In reality, the algorithm can easily become your new boss, controlling you very tightly, ensuring that you work quickly.

What about the pros and cons for employers?

Again, the main pro is the flexibility. You only have the workforce when you need it. The most significant drawback from the employer’s perspective would be that you lose a lot of goodwill from the employees. You lose the advantages of long-term cooperation. If you hire someone for a short time, it’s hard to know how good they really are.

What if you hire the same people again and again?

If you do that, why don’t you hire them on a longer-term contract?

Because that way the employers don’t have to pay them social benefits.

But then the workers would find another job. If you wanted someone really good, you would want to keep them. And that means that you would want to commit to them, to get them to commit to you.

It depends on who has a stronger hand in the labour market. If you have lots of unemployment, potential workers will be much more vulnerable.

We also see that in the numbers. During the recession, the gig economy was much bigger in structurally weak countries.

The future of work is not something that is happening to us but something that we have to shape and create ourselves

Isn’t the transformation of the labour market, including automation and robotisation, leading to a situation in which we will see even more of that vulnerability?

Are we moving towards a world of work where the gig economy is the main way of organising jobs? Whether we are or not depends on the choices we make. I strongly believe that the future of work is not something that is happening to us but something that we have to shape and create ourselves. I don’t believe in technological determinism. In terms of automation and robotisation, the epilogue of my book starts with John Maynard Keynes’ words. Nearly a century ago, Keynes wrote about how automation was going to destroy jobs and how people were going to work four hours a day. We still haven’t reached that point. Even though it is true that automation destroys some jobs, it also creates new jobs – jobs that we might not have thought about a few decades ago.

But the jobs destroyed are the ones done by low-skilled workers, while the jobs created favour the higher-skilled.

Here I have to mention what David Autor refers to as ‘Polanyi’s paradox’, which suggests that it is actually the low-skilled jobs that are harder to automate than the high-skilled jobs.[2] Many familiar tasks cannot currently be computerised because we don’t know ‘the rules.’ An intuitive job like cleaning the floor, where you sometimes have a cable lying around and sometimes don’t, is nearly impossible for a robot. The same applies for a delivery cyclist who has to dodge jaywalking pedestrians. Writing a contract, on the other hand, or making sure that its text is always spelled correctly is a very easy job for a robot, because many of these high-skilled jobs are fairly routine and predictable.

Unless of course you think of a lawyer, for example, who needs to find loopholes, because those are harder to automate than routine tasks.

I’m aware of at least one software competition where participants had to create a robot that would automatically fight parking tickets. It turned out that the robot had a much higher success rate in finding little loopholes and winning than a human lawyer.

One of the main problems with the gig economy is that its workers don’t have the same employment security as an employee. How can we improve that?

If someone is working under fairly tight control then, regardless of what his or her contract says, this person is an employee and should be protected. The Court of Justice of the EU, and lots of Member State courts, have long said that contractual terminology has little relevance if there is a completely different reality in fact. Under this doctrine of the primacy of facts, many people under the tight control of the algorithm would actually qualify as employees and therefore would fall under the scope of labour law.

What would enforcing labour law look like in practice?

If a person makes deliveries for many different people but the service is managed through the same platform, then the platform could be the perfect place to put that responsibility. In my book, I note that investors and shareholders would have to bear some of the costs of levelling the playing field and restoring platforms’ responsibility (not only towards their employees, but also regarding taxation and consumer rights). Levelling the playing field will undoubtedly cause disruption in the short term, but platform employers will have to face the full cost of their activities, from equipment and consumables, to wages and insurance premiums. The price of certain services will rise and some business models might turn out to be unsustainable altogether. But these steps have to be taken if we want the gig economy to work for everyone.

What if the gig workers work for many different platforms?

If they genuinely work for many platforms, then they may well be entrepreneurs. And real entrepreneurs are outside the scope of employment law. Employment law has historically drawn these distinctions. If a plumber, for example, uses a platform to grow their business and they set their own rate then they are an independent contractor. The problem cases are the ones where, based on the way they work, people look like employees but through tricks and contracts they are denied the status of employee. Those are the people the law can help when it examines facts rather than the contractual terms. Only those who actually enjoy the upsides of this way of work should be considered entrepreneurs.

If we look at Europe, do you see examples of good regulation in relation to the gig economy?

The European Commission is doing some potentially interesting things. The Social Pillar is still at an early stage and is not legally binding, of course. However, the recently launched proposal for a fundamentally reshaped Written Statement Directive has the potential to create more transparent working conditions by ensuring a level playing field, at least insofar as minimum working standards are concerned. That said, it’s equally important to remember that much of employment and labour law is still a matter for Member States, not least given the vast heterogeneity of labour market regulation across the EU. Finally, before rushing into new norms and standards, we should focus on the effective enforcement of existing standards – whether at the domestic or European level. In this regard, it will be fascinating to see how proposals for the European Labour Authority develop.

Regardless of how automation and robotisation unfold, we need to ensure decent working conditions across the entire labour market – including the gig economy. As long as people are working, we need labour law to ensure fundamental protection for workers and to protect consumers and markets, thereby encouraging innovation and competition.


[1] In December 2017, the European Court of Justice ruled that Uber should be regulated as ‘a service in the field of transport’ not as ‘an information society service’ as it had argued. See press release:

[2] Named after Michael Polanyi’s observation that “We know more than we can tell.” See Autor, D. (2014). Polanyi’s Paradox and the Shape of Employment Growth. National Bureau of Economic Research, Working Paper 20485.

This interview took place at the Corvinus University of Budapest, where Professor Prassl was lecturing as part of the university’s EFOP 362 project.

Cookies on our website allow us to deliver better content by enhancing our understanding of what pages are visited. Data from cookies is stored anonymously and only shared with analytics partners in an anonymised form.

Find out more about our use of cookies in our privacy policy.