National Data Reveal Europe’s Default Ecosystem Tourism Accounts Miss Half the Picture

When the European Union amended its environmental accounts regulation, member states took on a demanding new obligation: every year, they must report on the supply and use of seven ecosystem services, including nature-based tourism. To help countries comply, Eurostat offers a voluntary modelling tool called INCA, complete with ready-made datasets and default assumptions. A new study from Finland, published in the journal One Ecosystem, now shows just how costly those convenient defaults can be. When researchers replaced the pan-European input data with nationally sourced statistics and maps, the estimated volume of domestic nature-based tourism nearly tripled, and the map of where tourism happens changed beyond recognition.

The study, led by researchers compiling Finland’s nature-based tourism accounts for 2023, follows the System of Environmental-Economic Accounting framework, the international standard that underpins the EU reporting requirements. Under this framework, nature-based tourism is defined as overnight stays in accommodation attributable to ecosystem visits, meaning the contribution that the biophysical characteristics of ecosystems make to people’s direct, in-situ enjoyment of nature. The researchers ran the INCA tool twice: once with the default Eurostat datasets and once with a fully national set of inputs, including municipality-level travel statistics, spatial data on accommodation locations and adjusted proxies for recreational attractiveness.

The headline numbers are striking. The default approach attributed 8.1 million domestic overnight stays to nature-based tourism in Finland. The nationally adjusted approach attributed 21.5 million, nearly three times as many. The gap stems from two sources. First, the European travel statistics used by default count only commercial accommodation such as hotels, holiday rentals and camping grounds, excluding small establishments under twenty beds and all unpaid accommodation. Finnish statistics, drawn from representative population surveys, capture everything: 41.9 million domestic overnight stays in 2023, of which 46 percent occurred at own free-time residences, 14 percent at rental holiday homes and only 13 percent in hotels. With roughly half a million holiday homes in a country of 5.6 million people, each used on average 85 days per year, ignoring non-commercial accommodation means ignoring the backbone of Finnish nature tourism.

Second, the two approaches disagree on how much tourism is actually nature-based. The default method derives nature-attribution percentages from a pan-European Recreation Potential Map developed by the Joint Research Centre, assuming that the share of land with medium or high recreation potential in each region equals the ecosystem contribution. In Finland, this yields high percentages in densely populated southern regions and low ones in the sparsely populated north. The national approach instead uses a survey question on travel motives, letting respondents themselves indicate whether nature motivated their destination choice. Separate percentages were calculated for hotels, other paid accommodation and unpaid accommodation, yielding national averages of 21, 55 and 70 percent respectively. The highest values appeared in northern and eastern Finland, precisely where the default map underestimates.

The regional consequences are dramatic. In Etelä-Savo, a lake district dotted with free-time residences, the national accounting estimates over nine times more nature-based overnight stays than the default. In Lapland, the gap approaches a factor of twenty. Conversely, in Uusimaa, the capital region with its dense commercial tourism infrastructure, the default approach estimates about five times more than the national one. European statistics, the authors conclude, over-emphasise regions with commercial accommodation and overlook destinations where nature tourism flows through cottages and family cabins rather than hotel bookings.

The spatial modelling itself rests on three raster layers that the INCA tool multiplies into a recreation potential proxy: accessibility, landscape attractiveness and recreation facilities. The default accessibility layer relies on distance to settlements and main roads, marking large areas as poorly accessible. The national version instead uses the actual locations of hotels and free-time residences, distributing overnight stays within a 200-metre buffer around accommodations and within a 20-kilometre buffer around attractive destinations, mirroring how visitors actually behave: some recreation clusters around the cabin, while hiking and skiing spread across the wider landscape. For facilities, the researchers drew on the National Database of Sport Facilities, LIPAS, assigning weights to recreational areas, trails and facility points so that each infrastructure type contributes meaningfully despite differences in spatial extent. No comparable facility data exists at the European level, so the default datasets simply lack this layer.

The attractiveness layer received perhaps the most inventive national treatment. The researchers included waterbodies and shorelines with condition-based weights, adjusted coastline buffers to emphasise the water itself rather than the adjacent land, and added a novel element: forest clear-cutting. Drawing on Global Forest Watch data, forests that lost tree cover between 2003 and 2023 were classified as non-attractive for nature-based recreation, reflecting decades of Finnish research showing that large clear-cuts diminish the recreational experience in a country where freedom to roam grants access to all forests. This linkage between forest management and tourism supply illustrates how ecosystem accounts can connect to environmental condition indicators such as water quality and forestry practices, a connection the national nature recreation strategy and the tourism strategy both demand.

Validation supported the national approach. Comparing each recreation potential map against municipality-level estimates of actual nature-based overnight stays, the nationally adjusted map achieved a Spearman rank correlation of 0.54 with observed tourism intensity, against just 0.27 for the default map. The ecosystem-type breakdown shifted accordingly: in the default account, forests claimed 82 percent of supply with marine ecosystems excluded entirely, while the national account allocated 41 percent to forests, 38 percent to lakes, 9 percent to marine areas and meaningful shares to croplands, wetlands and rivers. The researchers also swapped the 100-metre pan-European land cover layer for Finland’s 20-metre national ecosystem type map, though this changed national-level shares by only one to two percentage points.

The implications reach beyond Finland. The authors note that Norway and Sweden share similar free-time residence cultures, and a Norwegian pilot had already flagged that INCA results require careful evaluation beyond minimum compliance. Because the default approach attributes nature-based tourism shares solely through the recreation potential map, trends unlinked to land-use change may go undetected; Finnish survey data show the nature-motivated share of domestic travel jumped from 53 percent in 2019 to 71 percent during the 2020 pandemic year, a shift the default method would likely miss. The researchers also flag open questions, including whether ecosystem-type-specific weighting should compensate for small but highly attractive ecosystems such as fells and coasts, and they point to mobility and social media data as promising avenues for refining assumptions.

The study’s conclusion is measured but firm: national data are essential to move beyond minimum compliance and produce ecosystem accounts robust enough for real policy decisions. The resulting accounts should be read as indicative representations of tourism flows rather than precise measurements, but they capture institutional realities, from freedom to roam to cottage culture, that pan-European defaults systematically miss. The INCA tool, its developers emphasise, is designed to be open and to accept national data, and the Finnish adjusted layers have been published in a public repository. For any country whose conditions diverge from European averages, the Finnish experience offers a clear lesson: the accounting framework is flexible, but the accounts are only as good as the data fed into them.

Subject of Research: National versus default input data in ecosystem accounting for nature-based tourism in Finland

Article Title: Improving nature-based tourism accounts: The importance of national data for robust ecosystem service estimates

Article References: Improving nature-based tourism accounts: The importance of national data for robust ecosystem service estimates. (n.d.). https://doi.org/10.3897/oneeco.11.e186675

Image Credits: AI Generated

DOI: 10.3897/oneeco.11.e186675

Keywords: ecosystem accounting, nature-based tourism, SEEA EA, INCA tool, Eurostat, Finland, ecosystem services, free-time residences, recreation potential, spatial modelling, EU regulation 691/2011, national statistics