Tourism in Italy: concentrated or widespread?

I'll start from the end, because the finding deserves it: Italy's tourism concentration index is, numbers in hand, virtually identical to Spain's — the textbook case of concentrated tourism in Europe. For years I argued the opposite, in conferences and posts, that Italy was the country of widespread tourism. I wasn't referring to the mainstream mantra of a thousand borghi (villages) and destinations, with heritage spread everywhere unlike Spain's big resorts concentrated in a handful of areas. I was looking instead at the territorial distribution of tourist flows. Specifically, I based my conclusion on the regional concentration of flows — more precisely the distribution of overnight stays at NUTS 2 level, which is how our "Regions" are identified in the official European Union nomenclature for classifying territorial subdivisions (NUTS stands for Nomenclature of Territorial Units for Statistics).

I was wrong (and in Note 1 at the bottom of the post I explain why). Let me get to the point. If we zoom in — that is, if we analyse the data at provincial level (NUTS 3 in EU nomenclature), the finest granularity available for a European comparison — the conclusion contradicts my thesis. Italian tourism is as concentrated as Spanish tourism — the textbook case of European tourism concentration — and more concentrated than French and German tourism. I must make amends, and that is why I find myself writing a long post to earn forgiveness. I promise to repay the time spent reading it with some interesting data.

How to measure the concentration of tourist flows

To measure how concentrated or distributed a phenomenon is, economists have used the Gini index for over a century — the same tool used to assess income inequality. How it works is simple: it ranges from 0 to 1.

• Zero indicates perfect distribution: applied to our case, it would mean that overnight stays are divided identically among all provinces.
• One indicates maximum concentration: almost all tourists cluster in a single province, leaving the rest of the country empty. The closer we get to 1, the more concentrated tourism is.

Two questions I'm sure you've already asked yourselves. First, why do I use presenze (overnight stays in accommodation establishments)? Simple. It's the most (territorially) disaggregated data we have for making this kind of comparison between European countries.

Second, why use an index like this instead of a more immediate figure, such as "how much weight do the top 5 destinations carry"? The strength of Gini is that it photographs the entire distribution, not just the top of the ranking. Two countries can have identical top-5 destinations by market share but completely different behaviour in the other 100. And it is precisely in the middle ("belly") and the "tail" of the ranking that the most interesting difference between Italy and Spain hides. But we'll come back to that.

Gini index of tourist flow distribution

Applying Gini to Italy's 107 provinces (Eurostat data, tourist overnight stays 2024), the result is 0.620. Spain's, calculated on its 59 provinces, is 0.629 — practically the same number. France's is 0.534, Germany's 0.610.

Chart 1. Gini index, concentration of overnight stays across provinces in Italy, France, Germany and Spain.

Gini index, concentration of overnight stays across provinces in Italy, France, Germany and Spain.

As the chart suggests, Italy and Spain are more concentrated than France (a gap of 0.086 points — wide and robust), but are essentially on a par with Germany (a gap of just 0.010, which is practically a statistical draw).

Two ways of being concentrated

Here comes the most interesting part. To understand it, imagine a school class and its overall performance. There are classes with two or three prodigy students who always score full marks, while the rest of the class performs around the average. Remove those two or three prodigies from the calculation and the class goes back to being a class like any other. This is the Italian model: the concentration is down to three provinces — Rome, Venice and Bolzano. Remove them from the count and the concentration level of the rest of the country drops to 0.536 — a number practically identical to France's starting figure (0.534). Italy, in other words, is "normal" everywhere except at three precise points. I wasn't all that wrong 😊 in the thesis I'm now disowning with this post.

Then there are different classes: not two or three isolated prodigies, but a broad group of eight or ten students all above average — a solid high-performing band, not isolated exceptions. Even if you remove the top three there, the class still outperforms the average, because the high level didn't depend on them alone. This is the Spanish model: remove the top three destinations (Mallorca, Tenerife, Barcelona) and the Gini barely drops, from 0.629 to 0.595, because behind them sits an entire block of islands and coastlines — Gran Canaria, Málaga, Alicante and others — that keeps concentration high even without the top three names. Spanish concentration is structural. Italian concentration is "spike-driven". In other words: in Italy the higher concentration compared to the norm (France, say) is not a characteristic of the territory — it is the effect of just three outlier "spikes."

The contribution of international tourism to concentration

The real Italian imbalance (and that of other countries) is almost entirely driven by international demand. Breaking the data down between domestic and foreign flows reveals a double reality: domestic tourism is well distributed (Gini 0.536, close to French levels), while foreign tourism is highly concentrated (Gini 0.726). Italy, therefore, does not suffer from a problem of structural imbalance: it simply has foreign demand that "squashes" itself onto a handful of provinces, while the rest of the country remains sustained and balanced by Italian tourists.

But beware: it would be a mistake to think this is a uniquely Italian peculiarity. Spain hides the same dynamic, in an even more extreme form. Spanish domestic tourism is the most distributed among the countries analysed (Gini 0.484), while its foreign tourism is absolutely the most concentrated (0.750). Spain's reputation as a "concentrated country" is therefore driven almost entirely by foreign visitors layered on top of a very homogeneous domestic market. Same story for Germany: isolating foreign tourists alone, the Gini index shoots up to 0.744 (nearly at Spanish levels), with Berlin and Munich alone capturing over a quarter of the entire foreign flow.

Let's broaden the view and cut to the root the question of the number of provinces — that is, how the administrative units identified by the NUTS3 acronym are distributed. I took the 10 countries with the most foreign overnight stays in 2024 and calculated a Gini index corrected to account for the number of provinces (I refer you to Note 2 for the explanation).

Country	Foreign overnight stays 2024	No. of provinces	Corrected Gini
Portugal	59.7 million	26	0.749
Croatia	85.0 million	21	0.747
Spain	322.2 million	59	0.736
Greece	128.2 million	52	0.727
Italy	253.9 million	107	0.701
Germany	79.3 million	360	0.699
Austria	94.0 million	35	0.686
Switzerland	27.7 million	26	0.645
Netherlands	61.5 million	40	0.620
France	140.6 million	101	0.614

What does the table tell us?

• The concentration of foreign tourism is a structural feature of tourism, at least when looking at the most visited countries. Among the "big four", only France "behaves" (partially) differently.
• Italy is confirmed in the middle of the ranking even among the big players: more concentrated than Germany, Austria, Switzerland, the Netherlands and France; less concentrated than Portugal, Croatia, Spain and Greece.

The flat belly of international tourism in Italy

We can picture the distribution of international tourists as a body divided into three parts:

• The head: a handful of phenomenon-destinations capable of attracting exceptional volumes;
• The belly: the intermediate band of established or emerging destinations;
• The tail: a multitude of small destinations with more modest numbers.

Read through this framework, Italy shows an unbalanced profile. It has a disproportionate head — those same three prodigies, Rome, Venice and Bolzano — and a very long tail of provinces with marginal volumes. But it has a thin belly, far thinner than one would expect. If we compare ourselves with Spanish tourism, what is missing is the central backbone. Looking at provinces capable of exceeding critical thresholds of foreign overnight stays (say, 10 million per year), Spain lines up around ten beyond its own giants — from Gran Canaria to Málaga, through to Alicante. Italy counts barely three or four before a sharp drop-off.

We don't lack champions and we don't lack capillary diffusion: what we lack is the middle tier.

How many destinations exceed a meaningful threshold of foreign tourism — Italy vs Spain, 2024

Chart 2. How many destinations exceed a meaningful threshold of foreign tourism — Italy vs Spain, 2024

To illustrate what I wrote above I made a chart showing, for five overnight-stay thresholds (from half a million to ten million per year), what percentage of the provinces in each country clears them. Note: these are not exclusive bands — a province that exceeds 10 million automatically appears in the lower thresholds too, because it satisfies all of them. This is why the chart should not be read as "distribution of provinces by band" but as a series of independent questions — "how many destinations have reached at least this scale?" — one per threshold.

The result: at every level, the share of Spanish provinces above the threshold is higher than Italy's, and the gap widens as the threshold rises. Above 10 million foreign overnight stays, Spain has 25% of its provinces; Italy has less than 10%.

Concentration is not a flaw — it is a normal phenomenon

Analysing this data carries an obvious risk: letting the message get through that "concentrated" is a synonym for "wrong." This is usually the "political" and journalistic reading. It isn't true, and it's essential to explain why. Concentration is not an anomaly of tourism nor an Italian distortion. It is the near-universal rule of the modern economy: finance clusters in a handful of stock exchanges, technology in a few metropolitan areas, fashion in a clutch of capitals. Tourism is no exception; it simply makes this dynamic more visible because we consume it by physically moving across the territory. In tourism, concentration is the outcome of three geographic and economic forces that intertwine and feed each other:

1. The geography of assets: the initial endowment

No country distributes tourism uniformly for the simple reason that physical geography, history and attractions are not distributed uniformly. Studies that have isolated the effect of natural or historical-archaeological endowment alone — the presence of sheltered coastlines, islands, cultural heritage — show that this resource by itself explains the origin of the main tourism poles. The concentration of supply does nothing more than mirror a physical and historical inequality that already exists from the outset. Rome and Venice did not become tourism poles because of a marketing campaign that worked better than others: they already were, in fact, before the very concept of a tourist destination existed.

2. The geography of markets: proximity and accessibility

Natural endowment alone is not enough: it must meet demand. This is where the position of a territory relative to tourist-generating markets comes into play. Geographic proximity or ease of connection with the major source markets (high-income countries) creates an enormous advantage. Low-cost flights and high-speed rail have partly helped peripheral tourist destinations, but transport economics follows market rules: an air or rail connection survives only if there is a sufficient volume of demand to guarantee high load factors. Source markets therefore reward territories that are closer or better positioned on major flow hubs, penalising peripheral ones regardless of their climatic, landscape or cultural value. It is no coincidence that Verona and Milan appear among the top five Italian destinations by foreign overnight stays: they are above all hubs, before they are destinations. This aspect intersects with the third force.

3. The district effect: agglomeration economies

Where initial endowment meets proximity and accessibility to markets, the critical mass that triggers the district effect is born. This is classic industrial logic: hotels, restaurants and services close to a key attraction and easily reachable draw enormous benefit from each other's presence. Proximity creates hyper-specialised suppliers, makes it easier to recruit skilled workers and accelerates the transfer of expertise between operators. It is the same mechanism as industrial districts applied to hospitality: a hotelier next to Piazza San Marco does not compete