Skip to content

The Price of Biodiversity

This page shows part of my research at Aarhus University. The research uses home sales to estimate what the ecological quality of nearby nature is worth to Danish homebuyers, as reflected in the prices they pay. I have found no earlier estimate of the price of an index of species and habitat quality in Danish property markets.

Work in progress. This page summarises preliminary findings from an unpublished working paper. The estimates are associations that should not be read as causal effects and may change before the paper is released. A revision in progress examines how much of the premium follows the type of land around a home, such as lakes and forest against farmland and built land. The map shows ecological-quality scores and does not show property valuations. Last updated November 2026.

An implicit price of zero

When a Danish public project such as a road or a building is appraised, its costs and benefits are weighed in money. Biodiversity usually enters that calculation at zero, because the guidance of the Ministry of Finance gives no unit value for it.

Denmark has a national map of ecological quality, the Biodiversitetskort, that scores each part of the country for its potential to support wild species. My research asks whether Danish homebuyers pay more to live near higher-quality nature and, if they do, how much more.

The main finding

The bioscore is the index of ecological quality in the national Biodiversitetskort and takes values from 0 to 19. Within 1 kilometre of the homes sold it averages 0.79, with 95 percent of the homes below 2.7. In my current research I match it to 741,867 Danish home sales from 2015 to 2025. The estimates below compare homes within the same parish (sogn), a local district with a median area of about 17 square kilometres.

about 2.5%

One bioscore point

When homes of the same type (detached house, terraced house or owner-occupied apartment) in the same parish are compared, one additional bioscore point within 1 kilometre of the home is associated with about 2.5 percent higher prices. At the median sale price of about €281,500 that is about €7,200. The 95 percent confidence interval, the range of values compatible with the data at the conventional standard of statistical evidence, is about €5,200 to €9,200.

about 1.5%

A typical difference inside a parish

One point is a large difference inside a parish, about 1.7 times the typical difference between homes of the same type there. That typical difference, one standard deviation of the score within those groups, is about 0.58 points. It is associated with about 1.5 percent higher prices, or about €4,200 at the same median price.

Both figures come from the same estimate and are associations. The estimate holds constant the characteristics of the home, its location and the greenness of its surroundings, measured from satellite images as the amount of green vegetation, and removes nationwide price movements quarter by quarter.

Ecological quality across Denmark

Average score within 1 kilometre of the homes sold in each parish, from 839,624 home sales of 2015 to 2025. Pick a layer below. A darker colour means a higher value.

Layer:

The full composite Biodiversitetskort score, the sum of the species score and the landscape score.

Use the arrow keys to move through parishes one by one, or use the Find a parish selector above the map. Press Enter to repeat the selected parish. The selected parish is announced in text below the map.

Bioscore (1 km mean)
Static map of Denmark shaded by average ecological-quality bioscore per parish. Coastal areas, north Zealand and the central Jutland lakes show the highest ecological quality. Dense cities and intensively farmed areas show the lowest.

The interactive version of this map needs JavaScript. The figure above shows the same data, the average ecological quality within 1 kilometre of the homes sold in each Danish parish.

How to read this. The colour shows the Biodiversitetskort score, a measure of habitat quality and of the recorded presence of red-listed species, averaged over the surroundings of the homes sold in each parish. Red-listed species are the species that the Danish Red List places in the categories regionally extinct, critically endangered, endangered, vulnerable, near threatened or data deficient. A higher score means greater potential for wildlife. The colours sort parishes into six groups by rank, from very low up to the top few percent, which means that they show how a parish compares with the rest of Denmark and do not show how far its score is from the top of the scale. The map is not a property valuation and gives no price for the nature near any single home. Parishes with fewer than 30 sales are left grey because an average over so few sales is too noisy to show. The Greenness layer uses a green scale and sorts parishes by satellite greenness in the same way.

The layers and the price estimates. The price estimates on this page use the composite score, except the estimate from repeated sales, which uses another index that is described in that section. The two parts of the score, the species score and the landscape score, are mapped so that a reader can see what the composite consists of in each parish. No separate price is reported for either part.

What each layer shows

The Biodiversitetskort score is the sum of two parts, a species score and a landscape score. The first three layers map the score and each of its parts. The fourth maps satellite greenness for comparison.

0 to 19 · species + landscape

Composite (bioscore)

The full Biodiversitetskort score, the sum of the species score and the landscape score. This is the measure behind the price estimates on this page, except the estimate from repeated sales.

0 to 9

Species (artsscore)

Weighted records of red-listed species, from observations in the Danish species databases and from habitats and ranges mapped by experts. Each record is weighted by the Red List category of the species, the certainty of its location, its age and the rarity of the species.

0 to 10

Landscape (proxyscore)

Indicators of landscape and habitat, each adding a point where its condition is met. They include mapped protected nature, a high share of forest and protected open nature in the surroundings, varied forest structure, low nitrogen deposition, few roads and ditches and a location within 1 kilometre of the coast.

satellite measure

Greenness (NDVI)

The amount of green vegetation seen from a satellite in the summer of 2021. It is shown for comparison and is not part of the Biodiversitetskort score.

View the data as a table

Find your parish

Pick a parish to see where it ranks in Denmark for ecological quality, measured by the composite score, and for greenness, measured by satellite NDVI.

The profile of a parish appears here when you choose it above or click it on the map.

Each parish with at least 30 sales gets one thin bar above, ranked from the lowest composite score on the left to the highest on the right and coloured by its band. The card is drawn in your browser and the button sends nothing. It downloads as a PNG image.

The parish profile needs JavaScript.

How far from the home the premium extends

The main estimate uses the score averaged over a circle of 1 kilometre around the home. Repeating the estimate with the score averaged over smaller and larger circles shows how far from the home this price difference, called the premium below, extends. For each circle from 100 metres to 1 kilometre, one bioscore point is associated with about 2.4 to 2.9 percent higher prices. Because the 95 percent confidence intervals of these estimates overlap, the data do not show where within that kilometre the association is strongest. The estimate is about 1.2 percent for the 2-kilometre circle and cannot be distinguished from zero for the 5-kilometre circle. I read this as the profile of an amenity that residents experience close to home, on walks and through what they see and hear from the house and the garden.

Circle:

Each point is a separate estimate on the same sales, with the same characteristics held constant, in which the bioscore is averaged over a circle of the stated radius. The shaded band is the 95 percent confidence interval. The six radii are placed at equal intervals along the axis, which is therefore not drawn to scale.

View the estimates as a table

Why the unadjusted association is larger

Because homes near high-quality nature differ from other homes in many respects, a simple comparison of their prices does not show what buyers pay for ecological quality. The three figures below differ in what is held constant.

  1. No adjustment about 11%
  2. Characteristics of the home and its location held constant, all homes compared about 8.6%
  3. The same characteristics held constant, homes of the same type in the same parish compared about 2.5%
Price difference associated with one bioscore point within 1 kilometre of the home.

The unadjusted figure is more than four times the within-parish figure. The characteristics held constant, called the controls below, account for 29 percent of the fall and the comparison within parishes for 71 percent. I attribute most of the gap to sorting between parishes, the tendency of buyers who value nature to settle in parishes with high ecological quality. Those parishes also differ in other respects, such as access to the coast, commuting distance and the housing stock.

Homes in the same parish can still differ in ways that the data do not record, a sea view being one example. A bound proposed by Oster (2019) gives a measure of how much such unobserved factors could change the estimate. It uses her benchmark assumption that a model which also contained the unobserved factors would explain 1.3 times as much of the variation in prices within parishes as the model with the observed controls does. Under that assumption, unobserved factors would need to be about twice as strongly related to ecological quality as the observed controls, and in the same direction, to remove the estimate.

The same homes across repeated sales

A comparison across homes cannot rule out sorting within a parish, since buyers who value nature may outbid others for the homes nearest to rich habitat. A second design therefore compares each home with itself across its repeated sales, which holds constant each characteristic of the home and its location that does not change between the sales. It covers about 103,000 sales of about 48,000 detached houses from 2015 to 2026.

Because the editions of the Biodiversitetskort differ in their data and methods, this design uses another index, the High Nature Value index. That index takes values from 0 to 13, is updated each year and ranks farmland and protected open habitat that depends on grazing or mowing by its estimated value for threatened and rare biodiversity. A one-point rise in the index within 250 metres of a home between two of its sales is associated with about 1.0 percent higher prices, about €2,800 at the median price. A weakness of this design is that part of the measured change in the index comes from new reporting of nature data or from updates of the map layers behind the index, without any change on the ground.

Measured per standard deviation of each index across homes, the two designs give about 2.5 and 2.0 percent, the second for detached houses only. They agree in direction and each is statistically significant.

How it is measured

The main estimate comes from a hedonic price model, a regression that explains the sale price of a home by the characteristics of the home and of its surroundings. The model is fitted to 741,867 sales of detached houses, terraced houses and owner-occupied apartments. The controls are the size, age and construction of the home, its energy label, its distance to the coast, to protected nature and to the nearest station, its noise exposure and the greenness of its surroundings. The model compares homes of the same type in the same parish, which gives 3,825 groups in 2,091 parishes, and removes nationwide price movements quarter by quarter.

The map above uses a larger set of 839,624 sales, because the model requires complete information on each control. Most of the sales that the model leaves out have no energy label at the date of sale. For each home the bioscore is averaged over a circle of 1 kilometre around it, using the 2021 edition of the Biodiversitetskort for each sale year.

The specification in technical terms

The dependent variable is the log sale price. The regressors are the mean bioscore within 1 kilometre, the mean NDVI within 1 kilometre, structural characteristics from the sale record and the building register, dummies for the energy classes, the log distances to the coast, to the nearest Natura 2000 site and to the nearest open station, and two noise variables. The model includes parish-by-property-type and year-quarter fixed effects, with standard errors clustered by parish. The bioscore coefficient is about 0.025, estimated on 741,867 sales. The percent figures on this page are 100 times the coefficient. The euro figures are calculated by applying exp(coefficient) − 1 to the median sale price, which is why €7,200 lies slightly above 2.5 percent of that price. My current research reports the amounts in Danish kroner. They are converted here at 7.46 kroner per euro, which gives a median sale price of about €281,500.

Greenness and ecological quality

Satellite greenness records the amount of green vegetation and contains no information on which species live in it. A mown lawn and a species-rich meadow can therefore look equally green from a satellite. Each adjusted estimate on this page holds greenness constant, which means that it compares homes whose surroundings are equally green.

Across parishes the rank correlation between greenness and ecological quality, a measure of how similarly the two order the parishes, is slightly negative, at about −0.13. In 61 percent of parishes the ranks on the two measures differ by more than 25 percentile points.

Satellite greenness
Map of Denmark shaded by satellite greenness. Parishes on Funen, on Lolland and Falster and in parts of eastern Jutland and Zealand show as deep green, while western Jutland, Bornholm and city centres are paler.
Ecological quality
The same map of Denmark shaded by the composite ecological-quality bioscore. Many of the parishes on Funen and on Lolland and Falster that looked green score low here, while coasts, dunes and the central-Jutland lake district score high.
The two maps shade the same country by greenness and by ecological quality. Many parishes on Funen and on Lolland and Falster look green from a satellite and score low for ecological quality. Lysabild and Skamby, for example, rank near the top for greenness and near the bottom for ecological quality. Parishes on dunes, heath and coast, such as Skagen and Rømø, show the reverse.

What the estimates measure and what they leave out

What they measure

  • The part of the value of ecological quality that buyers pay for through the price of a home, with ecological quality understood as the national index scores it.
  • An average association across Danish home sales, among homes of the same type in the same parish.
  • A value taken from market prices for a part of nature’s worth that Danish cost-benefit analysis usually leaves at zero.

What they leave out

  • A causal effect. Neither of the two designs identifies one, which is why both estimates are reported as associations.
  • The value people place on nature they do not live near, for example the value of knowing that species and habitats persist, and the value of services such as pollination, water filtration and carbon storage.
  • A property valuation, or a price for the nature near any single home.
  • A finished result. The paper is unpublished and the numbers may change before release.

Where this fits

This work is part of my research at Aarhus University on how to choose the right projects and make the best use of public resources, including how to measure and value nature. Its aim is to give cost-benefit analysis a documented value for one part of what nature near homes is worth, the part that buyers pay for through house prices.

A short summary of the working paper is listed with my other research. A link to the full paper will be added when it is released.

Data sources

Ecological quality: Biodiversitetskort (bioscore, 2021 edition), Danish Centre for Environment and Energy (DCE), Aarhus University.
Parish boundaries: DAGI, via Dataforsyningen, licensed CC BY 4.0.
Property transactions: Boliga.dk, 2015 to 2025, and 2015 to 2026 for the repeated sales.
Building characteristics: the national building register (BBR), via Datafordeler.
Energy labels: EMOData, Danish Energy Agency (Energistyrelsen).
Noise: strategic noise maps for 2017 and 2022, Danish Environmental Protection Agency (Miljøstyrelsen).
Train stations: © OpenStreetMap contributors, available under the Open Database License.
Coast and protected nature: Topografisk Landpolygon, Klimadatastyrelsen, via Dataforsyningen, and the Natura 2000 database of the European Environment Agency.
Greenness: Normalized Difference Vegetation Index from a Sentinel-2 composite of June to August 2021. Contains modified Copernicus Sentinel data 2021.
High Nature Value index: the annual High Nature Value maps used in allocating Danish agricultural subsidies.
Method reference: Oster, E. (2019). Unobservable Selection and Coefficient Stability: Theory and Evidence. Journal of Business & Economic Statistics, 37(2), 187–204.

Referring to this work: the working paper is titled “The Price of Biodiversity: Hedonic Evidence from Danish Housing Markets” and is not yet published. For press questions, use the contact page.

Get in touch about this research

I am glad to discuss the valuation of biodiversity and its place in environmental cost-benefit analysis.