The State of Freelancing in Spain 2026

Abstract

Between 7 and 20 September 2026, Remote Work Europe asked freelancers registered as autónomo in Spain what the experience is actually like. 407 people completed the survey, in English and Castellano.

They describe a system they have learned to operate within, but do not trust. Four in five say they could not manage the paperwork without paying someone, and three-quarters spend five hours a month or less on admin. The survey shows those two things together; it does not establish that one causes the other. What it does show is that the time spent is low and the anxiety is not. More than three-quarters live with a background anxiety about getting something wrong with Hacienda, the Agencia Tributaria, and just 15% believe an honest mistake would be treated fairly. That anxiety does not wear off with experience, either: it came through at the same level among people registered ten years as among those registered one.

These are not people passing through. Seven in ten both earn their best money from outside Spain and describe themselves as migrants building a life here, and six in ten of everyone who answered intend to stay indefinitely. Four in ten do all three. Whatever else this group is, it is not passing through.

Asked to change one thing, 219 of the 276 who answered named cost, tax, or the complexity around them. Thirty-five of them named the cuota itself, and eleven asked specifically for it to be tied to what you earn rather than charged flat.

Fewer than half would choose self-employment in Spain again, and only about one in eight would do it without hesitation.

This was a self-selected survey rather than a representative poll, so it describes the 407 people who answered and not Spanish autónom@s as a population. What it offers instead is a glimpse of what specific autónom@s actually think, in their own words and at length.

Background: who we asked, and why

Spain had 3,469,416 registered self-employed workers in August 2026, across the RETA and SETA regimes, and the number has been rising: 63,705 more than a year earlier, and 220,141 more than in August 2018 (Ministerio de Inclusión, Seguridad Social y Migraciones, 2 September 2026).

In September 2026 Remote Work Europe asked some of them questions that public statistics do not address: what the system costs you beyond the fees, whether it gets easier, and whether you would do it again knowing what you know now.

The people who answered are a specific group, and it is worth being precise about who they are, because the answers make more sense once you know. They have moved to Spain and registered as autónomo under the Régimen Especial de Trabajadores Autónomos, which makes them liable for a monthly Seguridad Social contribution, the cuota, whether or not they earned anything that month. Six in ten hold non-EU nationality, seven in ten work mainly for clients outside Spain, and six in ten intend to stay here indefinitely. So they are neither short-stay nomads nor Spanish nationals serving Spanish clients. A small handful have recently taken a baja from the system.

That matters for reading what follows, because it sets what these people are comparing against. Median gross annual earnings for employees in Spain were €24,497 in 2024, with a mean of €29,540, according to the Instituto Nacional de Estadística’s annual salary structure survey, which is the most recent year for which a median exists. Pay has risen since without transforming the picture: the INE’s quarterly labour cost survey puts ordinary wage cost per worker 3.6% higher in the first quarter of 2026 than a year earlier, roughly in step with consumer prices over the same period.

That is the domestic market our respondents are largely not selling into. 85% of them say their best-paid work comes from somewhere else.

We wanted their own words rather than only their ticks, so four questions were left open with no prompts and no options. 653 remarks were captured, a good number at considerable length, and some of it more candid than anything the tick-boxes could reveal. Those verbatims are the core of this report; the numbers are mostly here to say how common each thing was among the people who wrote them.

Methodology, and the sponsor’s role

Fieldwork ran from 09:00 CEST on Monday 7 September 2026 to 23:59 CEST on Sunday 20 September 2026, and 399 of the 407 completed responses were started inside that window. The other eight arrived either side of it: six during instrument testing in late August, one shortly before the launch and one just after the close. All eight are included in the 407. The six from testing answered earlier drafts of the questionnaire, which is why seven agreement statements in this report carry only a handful of answers and are not reported at all.

The fielded instrument was version es-2026.4, offered in English and Castellano, with respondents choosing their language at the start. Every question was optional, which is why the base moves from one figure to the next. The instrument was deliberately built to take under ten minutes.

The questionnaire carried 22 agreement statements on a five-point scale, three summary questions, one multi-select question on AI offering five positive options, one neutral and three negative with two mutually exclusive answers, fourteen classification and context questions, one question on who handles the respondent’s admin and tax, and four free-text boxes.

Recruitment ran through the Remote Work Europe mailing list and Facebook communities, 22 Spanish autónomo associations with Red Vértice forwarding to its member network, and Xolo Spain’s own customer list. A prize draw was offered as an incentive, organised by Xolo Spain, and 321 of the 407 respondents entered it.

What is reported. 408 people completed the survey; one response was excluded during cleaning, giving 407 valid completes. Because questions were optional, the base for each figure varies between 397 and 406, and we give it wherever it matters. We do not report a completion rate. The survey page writes a row whenever it is opened, so those rows include search-engine crawlers, link previews and our own page loads, and there is no honest way to separate them from people who looked and left.

The sponsor’s role, and what independence means here

Xolo Spain supported this research and distributed the survey to its customers. That relationship is real, and it shapes the data in ways worth stating plainly rather than burying in a footnote.

• Xolo Spain reviewed the questionnaire before fieldwork, for legal compliance. Seven agreement statements were cut from the draft instrument before it went live; that was our decision, made to bring the survey under ten minutes, and Xolo Spain had no part in choosing which ones went.
• 35% of respondents, 142 of the 406 who answered that question, named Xolo Spain as their provider. That figure has two causes, and both push it up. Xolo Spain mailed its own customers. And Remote Work Europe has been referring readers to Xolo Spain for more than six months, so some respondents are their customers because we sent them there. We cannot separate the two effects. It is not market share, and no comparison between providers appears anywhere in this report, because the way people were recruited would make any such comparison meaningless.
• Xolo Spain had no involvement in the analysis and no editorial control over this report. Remote Work Europe designed the instrument, ran the fieldwork, holds the data, and wrote what follows.
• Xolo Spain asked to see the report before publication and we agreed, as a courtesy to a sponsor. [Final sentence to be written after their review.]

This research was resourced internally by Remote Work Europe, which is a very small organisation. Xolo Spain’s provision of the prize draw made a real difference to how many people took part, and a survey of this length gets a great deal less attention without an incentive behind it. We are genuinely grateful for it.

A note on the instrument itself

Fifty-two people answered at least one question and then stopped. Twenty-eight of them, more than half, stopped inside the first block. That is a finding about our survey rather than about Spanish bureaucracy, and we are publishing it because it is useful to anyone building something similar. The opening block asked thirteen agreement statements one after another, which is a lot to put in front of someone who has just arrived. Next time the opening will be shorter, and the demanding material will come later.

Limitations: what this cannot tell you

This is a self-selected convenience sample. People chose to answer, and they reached the survey through channels that are not neutral: a newsletter about remote work, communities for migrants in Spain, trade associations, and one provider’s customer list. Remote Work Europe also makes recommendations to its readers about who to use for this kind of work, which shapes who its audience is before any survey begins. Nothing here supports a claim about Spanish autónom@s as a population. Throughout the report we write “among those who answered”, and we mean it literally.

For scale: if this had been a random sample of 400, which it is not, a simple proportion would still carry roughly ±5 percentage points at 95% confidence, and since it is not random, even that understates the true uncertainty. That is why no percentage here is given to a decimal place and why there is no significance language anywhere in the report.

1. The sample skews international. Six in ten respondents hold non-EU nationality and seven in ten work mainly for clients outside Spain. That is the population this research set out to reach, and it is not the profile of the autónomo register as a whole.
2. It skews English-speaking. 345 responses came in English and 62 in Castellano, and we have not weighted the data to compensate.
3. The provider mix reflects how people were recruited, not the market. Our own channels are not neutral on this either: Remote Work Europe refers readers to particular providers and has done so throughout the period these respondents registered. The methodology sets out what that means for the largest figure in the table.
4. There is no regional analysis. The regional spread reflects how the survey was distributed rather than anything about Spain, and the resulting cells are too small to analyse and small enough to risk identifying individuals.
5. Seven agreement statements were cut before fieldwork and are not reported. We cut them ourselves, to hold the survey under ten minutes.
6. The income question did not specify a basis. It asked for approximate annual income without saying gross or net, or turnover against profit, so respondents will have read it differently from one another.
7. Quotes are edited only for anonymity. Where someone’s combination of sector, region, nationality and time in Spain could identify them, we have removed the detail or paraphrased. Nothing has been edited for tone, and spelling and typing errors are left as they were written.
8. Where a quotation carries a label such as “registered 3–5 years”, that label was taken from the respondent’s own answers at the point of analysis, when the closed answers and the free text were still held together. The dataset we share externally has those two things deliberately separated, so the labels cannot be reconstructed from it. That is a privacy measure rather than an omission.

Some of this is people having a moan, and that is worth saying

We invited it. The survey was presented as a chance to say what you think about being autónomo, the free-text boxes carried no prompts and no word limit, and people used them accordingly. A number went well beyond the subject of the survey into broader political territory we had not asked about, and we have not quoted that material.

The pattern in the data is clear enough to describe. Two of the four open boxes appeared only for respondents who answered at one extreme or the other of a particular statement, and those boxes drew the longest answers by some distance: 13 of the 79 responses to one of them ran past 400 characters, against 15 of 276 in the general boxes. People who felt strongly wrote at roughly three times the rate of everyone else.

What a sample like this is actually good for

A single person describing something establishes that the experience was reported, and shows its shape. It does not establish how common it is. For a lot of what this survey asked, whether it happens at all had not been established, so an account from one anonymous respondent is a different kind of claim from a weak version of a statistic. When somebody writes that they pay €305 a month whether or not they earned anything, that is true of them, and it stays true however many people answered the survey.

So this is what the report offers: proof that these experiences exist, the range and texture of them, the patterns among the 407 people who answered, and a much clearer view of what is worth asking next time. What you will not find is a sentence claiming that Spanish autónom@s as a group believe anything at all.

The 22 statements, ranked

Respondents rated each statement from 1, strongly disagree, to 5, strongly agree. The bars show the proportion who agreed or strongly agreed, among those who answered that item.

• My best-paid work comes from outside Spain 85% n=406
• I consider myself a migrant building a life here, not a temporary expat 84% n=405
• I could not manage the paperwork without paying for help 80% n=405
• Everything in Spain is set up to reward employees more than autónomos 79% n=405
• I live with background anxiety about getting something wrong with Hacienda 77% n=405
• It is hard to find trustworthy, clear advice about tax and compliance 70% n=405
• The system makes me less willing to take risks with my business 69% n=404
• The flat cuota punishes people just getting started 66% n=405
• I find it hard to charge what I know I am worth 63% n=405
• What I pay bears little relation to what I will get back 62% n=405
• Dealing with officialdom in Spanish is a real barrier for me 57% n=404
• The lifestyle is worth earning less than I could elsewhere 54% n=405
• I understand exactly what I am paying in taxes and fees, and why 53% n=406
• Some days I seriously question whether staying registered is worth it 52% n=405
• Some months bureaucracy takes more energy than the actual work 43% n=405
• I would rather not understand the tax system, I want to outsource it 41% n=402
• Realistically I will have moved on before I benefit from what I pay in 32% n=405
• My contributions are a fair price for healthcare and a pension 32% n=405
• I feel generally optimistic about the future of freelancing in Europe 31% n=405
• I feel part of a community of people doing what I do 25% n=406
• Being autonomo has got easier, not harder, since I started 18% n=404
• I trust that an honest mistake would be treated fairly 15% n=405

Base varies by item because every question was optional. Seven further statements were cut before fieldwork and are not reported.

Do freelancers in Spain trust the tax system?

No. Fifteen per cent of those who answered believe an honest mistake would be treated fairly. Sixty-five per cent believe it would not.

That is one of the starkest negative results in the survey. It sits alongside a second finding that explains why it matters so much: 77% live with a background anxiety about getting something wrong with Hacienda. Seven in ten also say it is hard to find trustworthy, clear advice about tax and compliance in the first place.

So the picture is not of people who find the rules complicated. It is of people who have concluded that the system is not on their side, and who cannot easily find anybody authoritative to tell them otherwise.

The anxiety does not wear off. We looked at it by how long people had been registered, expecting it to fade with familiarity.

• Registered under 1 year 82% n=92
• 1 to 3 years 75% n=153
• 3 to 5 years 76% n=54
• 5 to 10 years 75% n=65
• 10 years or more 76% n=33

• Registered under 1 year 21% n=92
• 1 to 3 years 14% n=153
• 3 to 5 years 13% n=54
• 5 to 10 years 9% n=65
• 10 years or more 18% n=33

Ten years in, three-quarters still carry the anxiety. Trust did not improve with tenure in this sample, which is not what you would expect if unfamiliarity were the main cause. A single cross-section cannot rule out cohort effects.

The fine system. It feels like there is a punishment culture in Spain where an honest mistake can potentially lead to bankruptcy. Even if the mistake is not related to you but the accountant you employ or the Hacienda itself.Survey respondent, registered 3–5 years

I live in constant fear of the system. And I know that, realistically, I can not afford to continue like this.Survey respondent, registered 5–10 years

The distinction people draw is between their provider and the system, and it runs right through the free text. One respondent, explaining why they never question staying registered: “I have an asesoría that I trust and I’m not the type of person that worries about these kinds of things. I just trust everything works out.” Trusting whoever files your paperwork turns out to be quite separate from trusting what happens if something goes wrong, and it is the second that the anxiety attaches to.

The capacity to absorb that risk is not evenly spread, either. Holding money back against a bill you cannot predict is only an option if there is money to hold back:

I pay 305 euro monthly, regardless of the income. If I don’t have the money in my account on the day of the cuota being charged, there is no easy way to pay it even on the next day, and I am charges a retraso fee.Survey respondent

How much time does autónomo admin actually take?

Less than you would think. Three-quarters of those who answered spend five hours a month or less on admin and paperwork.

• Less than 2 hours 34% 135
• 2 to 5 hours 41% 165
• 6 to 10 hours 18% 71
• 11 to 20 hours 4% 16
• More than 20 hours 3% 11

That looks like a system working reasonably well, until you set it against the fact that 80% say they could not manage the paperwork without paying someone to help. The hours are low because most people have bought them away. Four in ten go further and say they would rather not understand the tax system at all.

What the money does not buy is peace of mind. The same people reporting five hours a month are the people reporting 77% background anxiety. So the burden here is not clerical. It is financial, and it is emotional, and paying a professional removes only the first kind.

We also tested whether things have improved. Spain has digitalised a great deal of this in the last decade, and you would expect the people who lived through it to have noticed.

• Registered under 1 year 24% n=91
• 1 to 3 years 16% n=154
• 3 to 5 years 22% n=54
• 5 to 10 years 13% n=64
• 10 years or more 21% n=33

Agreement never passes a quarter in any group. Nearly half of everyone ticked neither, which is its own answer.

We cannot say from this that nothing has improved, and there is a confound worth naming: people registered longer tend to run more complex businesses, so their admin may have grown for reasons unrelated to any digital service. What we can say is narrower and still worth saying. Agreement that the system has become easier stayed below a quarter in every tenure group, including the people registered longest. Seven of the 33 registered more than ten years did report an improvement, so it is not that nobody has noticed one. It is that most have not.

The perception is worth setting against what actually changed. Electronic signatures were given legal standing in Spain by Ley 59/2003, in force from March 2004, but handling affiliation, contributions and official notifications electronically only became compulsory for autónomos on 1 October 2018, under Orden ESS/214/2018.

At least 300 of the 406 people who answered that question registered after that date. For roughly three-quarters of this survey, the digital system is the only one they have ever known, so they are not comparing it with the paperwork of a decade ago; they have nothing to compare it with. The 33 respondents registered for more than ten years are the only group who lived through the change, and 21% of them say it has got easier.

Where do freelancers in Spain actually earn their money?

Somewhere else. 85% say their best-paid work comes from outside Spain, and 68% work mainly for international clients.

Seven in ten of those who answered are importing income and building a life. More than four in ten are doing both and intend to stay indefinitely.

That is worth stating clearly because the easy assumption about internationally-paid freelancers in Spain is that they are passing through. Among the people who answered this survey, they overwhelmingly are not. Six in ten said they regard Spain as home and intend to stay indefinitely.

• Staying indefinitely, this is home 49% n=243
• Long-term, but not forever 46% n=70
• Not sure yet 22% n=59
• A few years, then elsewhere 13% n=24

Which way the causation runs, we cannot tell from this. Disillusionment may be driving the exit plan, or the exit plan may make it easier to be disillusioned. The observation stands either way: the people least committed to staying are the ones the experience has already lost.

Why do people become autónomo in Spain?

We did not ask why people registered. Seventeen of the 259 who answered the “best thing” question volunteered that employment here did not pay enough.

I couldn’t work for an employer as the wages are so low. So I run my business and I enjoy itSurvey respondent, registered 10+ years

None. I’d rather be employed but I can’t find local work so I’m forced to being self-employed.Survey respondent

To make sense of that, it helps to know what the domestic market pays, and what has been happening to it.

Instituto Nacional de Estadística, gross annual earnings for employees

	2020	2024	Change
10th percentile	€9,587	€12,018	+25.4%
Median	€20,920	€24,497	+17.1%
Mean	€25,166	€29,540	+17.4%

The lowest-paid decile has gained around half again as much ground as the middle. In 2020 someone at the 10th percentile earned 45.8% of the median; by 2024 that was 49.1%. Successive rises in the statutory minimum wage are the mechanism usually credited.

That compression is good news for people at the bottom, and it is also part of why the middle of the Spanish market feels flat to a skilled freelancer. If you sell professional services domestically, the going rate has moved less than the floor beneath it.

Self-employment is not always a choice

Research by Raquel Carrasco (Universidad Carlos III de Madrid) and Virginia Hernanz (Universidad de Alcalá), published by the “la Caixa” Foundation’s Social Observatory in February 2021, put involuntary self-employment in Spain at 21.7%, against a European average of 16.9%. In the underlying 2017 EU Labour Force Survey, 12.4% of Spain’s self-employed said they became so because no job was found as an employee, and a further 1.4% because a former employer required it.

But most self-employment in Spain is not involuntary, and the same data says so. The two largest single reasons are continuing a family business (23.2%) and taking a suitable opportunity (22.5%). This evidence is from 2017; the EU has not repeated that survey module.

The question our survey could not answer

Economically dependent self-employment describes someone formally self-employed who relies on a single client, which shades into what is commonly called falso autónomo when it substitutes for employment. On this measure Spain is below the European average, not above it: 3.6% against 8.4%.

We cannot say how much of this exists among the people who answered, because we did not ask. The questionnaire asked where respondents’ clients are based, never how many they have. That is a gap in our instrument, not a finding, and it is the single clearest thing we would add next time.

Presenting the Labour Inspectorate’s Strategic Plan 2025–2027 on 2 October 2025, the Labour Minister Yolanda Díaz said the Inspectorate had, in that year to date, carried out more than 859,517 inspection actions and “detected 46,000 false self-employed workers”. The plan, published in the Boletín Oficial del Estado, makes false self-employment a named workstream and commits to at least 500 additional inspectors.

Its cross-border work sits in a separate part of the document, built around posted workers, letterbox companies and the European Labour Authority. A person resident in Spain, registered as autónomo, invoicing a single company with no presence in Spain at all, is not a scenario the plan addresses in the text we read. We are not saying the framework cannot reach it. We are saying this report has not established how the authorities apply those frameworks to a Spain-resident freelancer whose only client is overseas.

Do freelancers in Spain feel part of a community?

Mostly not. A quarter feel part of a community of people doing what they do. Nearly half actively do not.

• Staying indefinitely, this is home 26% n=247
• Long-term, but not forever 25% n=72
• Not sure yet 22% n=59
• A few years, then elsewhere 21% n=24

Flat. People who have bought a house, married locally and intend to die here report feeling professionally isolated at essentially the same rate as people leaving in two years.

That removes the comfortable explanation. Community agreement was much the same across every horizon group, from 21% to 26%, so how long somebody intends to stay explains very little of it. What does explain it, this survey does not test. Working alone, distant clients and the language barrier that 57% report are all plausible and none of them were measured against this.

The statement asked whether people feel part of “a community of people doing what I do”, which is a question about professional peers. We did not ask how rooted anybody is in their own town, family or friendships, and somebody with a full personal life and no professional peer group would answer this exactly as somebody genuinely isolated would.

People answered it as though we had asked. Given a statement about professional peers, respondents wrote back about friendships, cities and whether their life here feels like the one they wanted. Freelancers often do not draw a firm line between working life and personal life, and these answers suggest the two are being felt as one thing rather than two.

Making local friends has been SO hard and I don’t feel like I am part of any community in the city I live in, something that in the UK I felt very strongly.Survey respondent, registered 10+ years

What has AI done to freelance income in Spain?

It has made people faster without making them richer. The most common answer, by a wide margin, is speed with no change in income.

• I work faster, but my income has not changed 43% 170
• It made the admin and paperwork side easier 31% 125
• It improved the quality of what I deliver 29% 117
• I work faster and earn more 27% 107
• Clients expect more work for the same price 22% 87
• It replaced tasks I used to pay someone else to do 18% 73
• It reduced demand for my services 18% 71
• It opened up new services or new clients 16% 63
• Little or no impact 10% 39
• It reduced my rates or income 9% 36
• I do not use AI tools 8% 30

Multi-select, so these overlap and do not sum to 100%. Two options were mutually exclusive and cleared any other selection.

The question allowed more than one answer, so these groups overlap rather than oppose each other. 43% report working faster without earning more, 27% report working faster and earning more, and 22% say clients now expect more work for the same price. Forty-eight people, 12% of those who answered, chose both the first and the third.

The people who say AI has not touched them

Thirty respondents told us they do not use AI tools, and thirty-nine said it had made little or no difference either way. Taken at face value, that is around one in six for whom none of this applies.

It is worth setting against what is happening underneath. Eurostat puts AI use among Spanish firms with fewer than ten employees at 10.1% in 2025, up from 5.7% a year earlier. Large enterprises are at 57.5%. Those figures count businesses that know they have adopted something. They do not count the AI that arrives inside software people already pay for: the scheduling tool, the accounting platform, the inbox, the phone.

There is a pattern to how technologies settle, and AI is partway through it. They arrive as products people choose and pay attention to, and they end up as infrastructure people simply use, at which point the name stops being mentioned at all. Spam filtering, card fraud detection, autocorrect and the ranking of a social feed were all once described as artificial intelligence. Nobody calls them that now, and nobody chooses them either. The question “do you use AI?” only works while the answer is still a purchase.

Our own data shows what that looks like in practice, and it is not what we expected. The people who say AI has not touched them are the least likely in the survey to be using an online platform, and the most likely to be using a human gestoría: 25% against 39% across all respondents. Of the 69, four do their own admin. Fifty-four agree they could not manage the paperwork without paying somebody, 51 spend five hours a month or less on it, and 35 use a gestoría against 41% across all respondents. This is the most thoroughly delegated group in the survey. What that delegation means for how closely they follow their own compliance is not something we asked.

That is why the strongest signal in our own data is not about tools at all. It is that 43% work faster without earning more, and 22% say clients now expect more work for the same price. Whether people who do not use AI are also being repriced by it is a real question, and one this instrument cannot answer: we made “I do not use AI” exclusive of the impact options, so non-users were never asked about market effects. That is a design lesson for the next wave.

In the interests of consistency, this report is one of the examples. Remote Work Europe is a very small operation, and the analysis behind these findings, counting 407 responses across 29 questions, cross-tabbing them and grouping 653 pieces of free text, was done with AI assistance. Ten years ago a team would have been needed, or it would not have happened. The editorial judgements, the decisions about what could and could not be claimed, and the words, are ours.

What would autónomos in Spain change about the system?

Make the cuota proportional to income. That single request dominates everything else people asked for.

276 people answered this question, and 219 of them, four in five, named cost, tax or bureaucratic complexity.

make it simpler⁵⁰ deductions are guesswork¹⁴ proportional to income²⁵ the cuota itself¹²⁴ tax burden¹¹⁸ tarifa plana too short¹⁸ IVA threshold⁸ fear of fines¹² paro and sick pay²⁴ not needing a gestor⁸ paying when not earning²¹ quarterly prepayment⁹

Themes are coded, not raw word frequency, and Spanish and English phrasings of the same theme are merged. Counts are floors: pattern matching catches explicit phrasings and misses implied ones. Answers code to more than one theme, so they do not sum to 276.

• Cost: cuota, social security, monthly charges 124
• Tax, IRPF, IVA, Hacienda 118
• Simplify it, make it clearer 50
• Make the cuota proportional to income 25
• Paro, sick pay, parental leave, employee-equivalent rights 24
• Extend the tarifa plana, help in the first years 18
• An IVA threshold for small businesses 8
• Less dependence on a gestor 8

Overlapping categories. Read as at least this many, not exactly.

Autónomo fee just needs to be a percentage of profit rather than these bands. And should be a far lower fee to reflect the risk and lack of benefits that autónomos take in Spain compared to employees.Survey respondent, registered under a year

La cuota de autónomo debería ir en función de tus ingresos facturados.Survey respondent, registered 1–3 years

Four responses out of 276 asked for the cuota to be abolished outright. Almost everything else was about the shape of the charge rather than the principle of paying it. What people object to is a fixed monthly sum that arrives whether or not they earned anything:

Paying hundreds of euros for the cuota during months without any real income isn’t a social safety net, it’s a financial penalty for being autonomo.Survey respondent

Pago para trabajar.Survey respondent

And the best thing about it?

259 people answered that question too.

lifestyle and weather⁴³ international clients²² escaping low salaries¹⁷ no Spanish boss¹¹ time with family²⁶ nothing²⁰ nothing specific to Spain¹⁴ choosing my clients¹⁵ freedom and flexibility⁹³ the visa⁹

Same coding rules. Counts are floors.

• Freedom, flexibility, setting my own hours 93
• Lifestyle, weather, quality of life 43
• Nothing good about it 12
• Nothing specific to Spain 7

Twelve said there was nothing good about it, two of them in a single word. Seven others made a sharper point: they named something good, and then said it had nothing to do with Spain.

Spain itself isn’t why freelancing has plus sidesSurvey respondent

Todo lo bueno de trabajar por cuenta propia no está relacionado con el hecho de estar en España.Survey respondent

Would they do it again?

Fewer than half would. Only about one in eight would do it without hesitating.

• Definitely yes 13% 51
• Probably yes 29% 117
• I am not sure 31% 123
• Probably not 17% 67
• Definitely not 10% 41

• Very easy 3% 13
• Quite easy 17% 68
• Neither easy nor difficult 29% 116
• Quite difficult 41% 165
• Very difficult 9% 37

42% yes. 31% not sure. 27% no. That is the number this report exists to sit with. The findings above are the attitudes most strongly associated with it in this sample, which is not the same as showing they caused it: a trust deficit that experience does not cure, a cost that arrives whether or not the work does, a professional isolation that permanence does not fix, and a domestic market most of them have given up selling into.

It is not the whole picture, and it would be dishonest to end there. Two in five would do it again, and a proportion of them are emphatic about why:

I am now retired and have been treated fairly by the system in Spain.Survey respondent

I was scared of being autónoma in the first 2 years but if I hadn’t stuck it out until the 3rd year I would have missed out on landing a huge client that changed my life. So the struggle in the beginning was so worth it.Survey respondent

I don’t mind paying taxes when I can see what they fund. Living in Spain, I benefit from public healthcare, infrastructure, public services, safety, education, and countless things that make everyday life easier and more stable.Survey respondent

We did not ask whether people trust their provider, so we cannot say that finding a trusted one reduces the fear. What the data does show is that buying help is not obviously doing it: anxiety sat between 69% and 78% in every provider group, and the lowest of the four readings belonged to the small number of people who file everything themselves. Whatever reduces this, the survey does not show that paying somebody is it.

Where this research goes next

Running this once produced a map of what to measure properly. Three things follow from it.

The questions we would add

How many clients do you have? We asked where respondents’ clients are based and never how many there are. That single question separates a freelancer with a portfolio from someone economically dependent on one payer.

Income on a stated basis, naming gross or net and turnover or profit, so the answers are comparable with published earnings data. Have you changed provider, and did anything improve? We measured where people are now, not movement. And region from the first day of fieldwork.

What we expect to move

• The AI gap. 83% of our respondents report AI affecting their work in some way. Eurostat measures something different – whether firms have adopted specified AI technologies – so the two describe different populations answering different questions, and should not be read against each other. How individual freelancers experience AI arriving rather than choosing it is not something either measure currently captures.
• Enforcement. The Labour Inspectorate’s plan runs to 2027 with 500 additional inspectors. Detection should rise. Whether that changes how autónom@s feel about the tax authority is the more interesting question.
• The cuota. If policy moves, these answers are the baseline for asking whether it changed how anybody feels.
• Trust. Fifteen per cent, and flat across every tenure band. A measure that does not shift with familiarity is structural, and structural measures can move when a system changes.
• The undecided. Where that 31% has landed in twelve months is the most informative single number we could collect.

The comparison worth making

A single-country study cannot tell you which of its findings are about Spain and which are about self-employment. Remote Work Europe covers more than thirty European jurisdictions, and running the same instrument in two or three of them would separate the two. Estonia would test whether the administrative findings are about Spanish systems or about self-employment itself, since filing there is fully digital by default. Portugal has a comparable contribution structure. Ireland and the Netherlands run different social insurance models against a similar cost of living.

The trust finding is the one we would most want to test elsewhere. If 15% turns out to be a Spanish number, it describes a policy problem with a national fix. If it turns out to be a European number, it describes something considerably larger. A repeated cross-country measure of perceived fair treatment would settle which, and we are not aware of one.

Acknowledgements

This research was supported by Xolo Spain, who also distributed the survey to their customers. The independence arrangements are set out in the methodology.

Thank you to the people and organisations who put the survey in front of their own communities and genuinely moved the numbers: Live Galicia, Valencia Nomads Hub, and Red Vértice, which forwarded it to its member network.

[Diana to be credited here in her own right, with her own link.]

And to the 407 people who answered, and the 297 of them who also wrote to us in their own words. The survey was offered as a chance to say what being autónom@ in Spain is actually like. People took it seriously, and this report exists because they did.

Sources

• Registered self-employed in Spain — Ministerio de Inclusión, Seguridad Social y Migraciones, monthly affiliation release, August 2026. lamoncloa.gob.es
• Employee earnings, median and percentiles — INE, Encuesta Anual de Estructura Salarial, table 28191, 2024. ine.es/jaxiT3/Tabla.htm?t=28191
• Wage growth to Q1 2026 — INE, Encuesta Trimestral de Coste Laboral, table 6038, series ETCL1526. ine.es/jaxiT3/Tabla.htm?t=6038
• Involuntary and dependent self-employment — Raquel Carrasco and Virginia Hernanz, “la Caixa” Foundation Social Observatory, February 2021, on 2017 EU-LFS data excluding agriculture. elobservatoriosocial.fundacionlacaixa.org
• Reasons for becoming self-employed — Eurostat lfso_17sereas, 2017 EU-LFS ad-hoc module, Spain. ec.europa.eu/eurostat
• Enforcement figures — La Moncloa press release, 2 October 2025, quoting the Labour Minister at the ITSS Strategic Plan launch. lamoncloa.gob.es
• Strategic Plan 2025–2027 — BOE-A-2025-18078. boe.es
• Electronic filing obligation — Ley 59/2003 (boe.es, since replaced by Ley 6/2020) and Orden ESS/214/2018, BOE-A-2018-3080 (boe.es)
• Enterprise AI adoption — Eurostat isoc_eb_ai, indicator E_AI_TANY, updated 15 June 2026. ec.europa.eu/eurostat

Every figure above was re-verified at primary source on 25 September 2026, and every link was checked to resolve on that date.