Skip to main content

Advertisement

SpringerLink
Log in

Wind power deployment and the impact of spatial planning policies

  • Published:
Environmental and Resource Economics Aims and scope Submit manuscript

A Correction to this article was published on 01 February 2024

This article has been updated

Abstract

Spatio-environmental externalities of renewable energy deployment are mainly managed through spatial planning policies, like regional expansion goals, zoning designated areas, or setback distances. We provide a quantitative analysis of how effectively spatial planning policies can steer RES deployment, using the example of onshore wind power expansion in Germany. Based on a novel georeferenced dataset of wind turbines and spatial planning policies, we use a dynamic panel data model to explain yearly additions in wind power capacities. Most importantly, we find a strong positive impact of zoning specific land areas for wind power deployment. An additional square kilometer of designated area leads to an increase of 4.6% of yearly capacity additions per county. Not only the amount of designated area matters, but also the size and shape of each individual designated area. Small and elongated areas are, on average, associated with more wind power expansion than large and compact areas. Moreover, we find that in states with an expansion goal, capacity additions are 2.6% higher. In contrast, increasing the setback distance between turbine sites and settlements by 100 m is associated with reductions of yearly capacity additions by about 3.1%. Our findings show that policymakers can resort to spatial planning instruments in order to effectively arrange wind power deployment with other land uses.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

Change history

Notes

  1. Stede et al. (2020) show that construction permits for wind turbines declined in response to a larger setback distance. This decline reflected the extent to which available land for wind power deployment was reduced due to the the increase in setback distance.

  2. In their study on wind power expansion in Sweden, Ek et al. (2013) include a dummy variable indicating whether municipalities have areas classified as national interest for wind power (NIAW). They find a positive effect for this NIAW classification. Since the NIAW classification is their only proxy for good wind resources the positive effect cannot be attributed to potential facilitation of planning and permit processes.

  3. For an extensive description of the spatial planning system in Germany see Pahl-Weber and Henckel (2008).

  4. Of course, there are many further ways how state governments can regulate wind power deployment, e.g. by means of the state planning act, the state building law, or their nature conservation law. Equally, there are further important policy decisions taken by state governments, e.g. the specification of setback distances between wind turbines and aeries of threatened bird species.

  5. As expansion goals we refer to specified target marks that may have the legal form of goals or principles of spatial planning (“Ziele oder Grundsätze der Raumordnung”), or the form of sectoral planning decisions (“Fachplanung”) or administrative provisions (“Verwaltungsvorschrift”).

  6. Setback distance rules can be defined with respect to different aspects. They can refer to the height of wind turbines or the type of settlement.

  7. In general, regional and local planning authorities have to take into account the respective others’ zoning decision. The Federal Spatial Planning Act prescribes this ’mutual feedback principle’ in order to ensure consistent spatial planning.

  8. Precisely, the Federal Spatial Planning Act (FSPA, “Raumordnungsgesetz”) sets three types of designated areas: priority areas, restricted areas, and suitable areas (§7 Para. 3 FSPA). Utilizing one of these types of designated areas or a combination of them regional planning authorities de facto implement what we denote as exclusive and non-exclusive designated areas.

  9. This is a valid assumption since most sites are allocated to wind developers in a first-come, first-served manner. Commonly, wind developers attempt to secure suitable locations as early as possible, e.g. by contacting land owners and concluding user contracts. Over time, land owners (most of them are farmers) professionalized in leasing their land, such that price competition over land use rights emerged among wind developers. In model terms, assuming a representative wind developer per county versus perfect competition among many wind developers is equivalent as long as the representative firm optimizes the net present value of those profits that originate from current but not future decisions.

  10. Precisely, \(e_{i,t}\) represents the expected sum of electricity produced from wind power capacities over their life time.

  11. See Appendix H for more information on the details of the federal RES support scheme.

  12. Again, this is discounted total cost over the operating time of a wind turbine. However, from a wind developer’s perspective 90–95 % of total costs of wind power projects are upfront, e.g. turbine price, project planning, etc. (Wallasch et al. 2015).

  13. Site productivity \(w_{i,t}\) has unit h. Site productivity indicates the amount of full-load hours, i.e. the period of time the wind turbine is operating at full capacity.

  14. As both of them are latent variables we cannot observe the effect of spatial planning policies on them directly. \(y^*_{i,t}\) is at least to a certain extent private information of the wind developer, and \(\bar{y}_{i,t}\) is not clearly definable because technical and legal restrictions for potential wind turbine sites are not conclusively assessable. Still, on the basis of our model we can hypothesize which effects on \(y^*_{i,t}\) and \(\bar{y}_{i,t}\) we are expecting, and which overall effect on \(\hat{y}_{i,t}\) is expected.

  15. Deployment cost comprise construction as well as operation costs linked to deploying wind power capacities. \(c_{i,t}\) has unit € /MW.

  16. This rests upon the realistic assumption that \(s + w \frac{\partial s}{\partial w} > 0\), which means that the RES support scheme is designed such that the marginal revenue of an additional unit of wind power capacity \(s_{i,t} w_{i,t}\) increases with site productivity. This is true for the German RES support scheme, see Appendix H.

  17. The project length can be divided into the earlier ’time-to-plan’ (TTP) period and the later ’time-to-build’ (TTB) part of wind power projects. With regard to wind power projects the TTP period accounts for most of the project length. For our investigation period, the maximum length of the TTP phase is about four years (FA Wind 2015). In general, the TTP period plays an important role for commercial construction projects (Millar et al. 2016).

  18. Nowadays, wind developers have to provide e.g. formal environmental impact assessment, run public participation procedures, or order expert reports concerning monument protection, visual axes, etc. Furthermore, data on the realization period of wind power projects between 2000 and 2016 underpins that this phase of wind power projects was considerably shorter compared to today. The realization period spans from the day the building permission was issued to the day of inital operation. Between 2000 and 2016 average realization period was among ten to 15 months (Fachagentur Windenergie an Land 2021). Exceptional cases are the years 2008 and 2009 when average realization period was above 20 months.

  19. Of course, the argument could also work the other way around.

  20. We assume that population density: and GDP per capita are strictly exogenous, while green votes and all federal and state policy variables are defined as predetermined variables. Predetermined variables are not strictly exogenous as they are potentially correlated with past and current error terms (Roodman 2009).

  21. Errors of the first-differenced regression model should have negative first-order serial correlation and zero second-order serial correlation to reasonably assume no serial correlation of error terms in levels (Kiviet 2020). We use postestimation statistics of the STATA-command xtdpdgmm to implement AR(1) and AR(2), see Kripfganz (2019).

  22. When estimating FE and RE models the within estimator is always inconsistent since the demeaned regressor is correlated with the error term (Cameron and Trivedi 2005).

  23. We treat county mergers that took place between 2000 and 2016 as if they had occurred at the beginning of our period of investigation. Only 29 of 401 counties (7.23 %) were restructured due to administrative reforms, almost all of them (28 counties) are in Mecklenburg-Western Pomerania, Saxony, and Saxony-Anhalt.

  24. The data set by Manske et al. (2022) is open access and provides the most accurate picture of the spatial and temporal distribution of wind power expansion in Germany.

  25. The variable total designated areas includes all designated areas independent from the type of planning—exclusive or non-exclusive planning. However, by interacting the variables exclusive planning and total designated areas in our regression analysis, we account for potential differences of the effect of zoning designated areas under one or the other type of planning.

  26. Usually, the state level decides which type of planning is in place. Though state governments prescribe the type of planning, we subsumed the variable exclusive planning under regional policies because it describes the actions by regional planning authorities.

  27. A more detailed number of technical potential for wind power deployment is provided by Masurowski et al. (2016). His data on technical potential even takes into account the legal lower bound of setback distances. However, his data also incorporates some state rules which in our analysis comes within the policy variables. Using the data by Masurowski et al. (2016) results does not change.

  28. Table 7 in Appendix A presents the correlation matrix.

  29. Table 11 in Appendix C presents regression results when using wind power additions measured in MW respectively MW/km\(^2\) as the dependent variable.

  30. If we assume the true value is \(\lambda = 0.35\), then the FE estimate \(\hat{\lambda }^{FE}\) approximately deviates by \(- \frac{1+\lambda }{T-1} = - \frac{1+0.35}{12-1} = -0.13\) as \(N \rightarrow \infty\) (Nickell 1981).

  31. We apply a log-linear model (see Sect. 3) such that a one-unit change in a regressor variable induces a 100\(\times \beta\)% change in the (not transformed) dependent variable.

  32. We find a similar magnitude of the effect of zoning an additional 1 km\(^2\) of designated area per county when using the change of absolute capacities (MW) and applying the conditional FE-Poisson estimator. The estimated coefficient indicates a rise of 2.8% (see Table 11). Likewise, when using the change of capacity density (MW/km\(^2\)) as the dependent variable, the estimated effect of an increase in the share of designated area by 1 percentage point is an increase in yearly capacity additions per km\(^2\) by 28.1% (see Table 11). By the regression results of our main specification we need to add 6.1 km\(^2\) to achieve a 28.1% rise in yearly capacity additions. Given the average county area of 890.8 km\(^2\) adding 6.1 km\(^2\) of designated area corresponds to 0.7 percentage points.

  33. As depicted by Fig. 10f, the impact response to the implementation of exclusive planning is rather delayed by more than one year. Since we interact exclusive planning with total designated area and avg. capacity density we include the first lag of exclusive planning also in our main model specification. Lagging exclusive planning by more years does not change its estimated coefficient neither the other estimates.

  34. In Germany, regional planning authorities follow a planning procedure by which, first, those areas have to be excluded from the pool of possible designated areas which meet so called “strict taboo criteria” (e.g. settlement areas) and, second, further areas can be excluded which meet “soft taboo criteria”. Depending on regional socio-geographic conditions, this leaves differing leeway for regional planning authorities. Also, there can be vague requirements for the zoning of designated areas that need to be considered by regional planning authorities. These intend to ensure that planning authorities select sites where the deployment of wind turbines is technically and economically realizable. However, the very reason and need for spatial planning in the first place, namely coordinating and trading-off multiple public and private interests that demand for land use, must leave—at least to some extent—discretionary leeway to the regional planning authority.

  35. As explained by means of the theoretical example in Appendix F, ceteris paribus, changing square-shaped areas to form narrow rectangles changes the value of the variable average shape of individual designated areas by \(\frac{1}{25} - \frac{1}{16} = -\frac{9}{400}\). Hence, the effect on yearly capacity additions is equal to 93.4 % = \(-\)41.5% \(\times\) \(-\frac{9}{400}\)

  36. We control for the profitability of the site by including total remuneration and only consider the capacity density five years after a designated areas came into effect. Five years after the zoning of areas available sites are usually used up. See also Fig. 9.

  37. We do not track whether regional plans allow that the rotor blades of wind turbines may cross the boundaries of designated areas or not. If this is forbidden by the regional planning authority, the effect should be less negative respectively larger.

  38. Equally, the positive estimate of technical potential land from the RE model reflects this.

  39. The negative effect of the overall density of existing wind turbines further suggests that also turbine sites outside from designated areas are used up. It should be noted that we have no data on designated areas that are specified at the local level. Therefore, wind turbines which are constructed outside of regionally planned designated areas may lie within locally planned designated areas.

  40. The weak positive correlation of total designated areas and wind power potential may reflect this (see Table 7).

  41. Optimization studies usually apply a siting algorithm that ensures the optimal exploitation of available land in terms of maximally installable wind power capacities.

  42. Winikoff (2021) highlights the role of landowners in influencing the deployment of wind power. They show that in areas with fragmented landownership less wind turbines are installed than in areas with concentrated landownership.

  43. For their analysis Bons et al. 2019, p.66 f.) assume an average value of capacity density of 25.64 MW/km\(^2\) for Germany.

  44. Enevoldsen and Jacobson (2021) study existing wind farms and draw geometric areas around wind turbine clusters to derive estimates of capacity density with mean values of 19.8 MW/km\(^2\) for operational wind farms in Europe, and 20.5 MW/km\(^2\) outside of Europe. For European onshore wind farms they find a median value of 13.9 MW/km\(^2\).

  45. The variation of capacity density is also illustrated by the development of the average capacity density across German states, see Fig. 19 in Appendix G.

  46. The lack of county-level data leaves more work for further research. For example, an interesting question is whether exclusive or non-exclusive planning at the regional level leads to more total designated areas. Since under exclusive planning all relevant matters are weighed at the regional level and conflictual issues are resolved by the regional planning authority, allegedly there might be less blocking opportunities for local planning authorities that intend to avert wind power deployment on their territory.

References

Download references

Acknowledgements

We thank two anonymous reviewers for their elaborate comments which greatly improved our paper. Also, we are especially grateful for the valuable exchange with Sven Kripfganz, Raffael Stretz and participants at the EAERE 2022, the LEEP seminar, the MultiplEE Workshop 2021, the 11th Bamberg–Halle–Jena–Leipzig Workshop on Empirical Microeconomics, and the Energy Group at the Helmholtz Centre for Environmental Research UFZ. The current work has been funded by the German Federal Ministry of Education and Research under Grant 01UU1703.

Author information

Authors and Affiliations

  1. Institute for Infrastructure and Resource Management, Leipzig University, Ritterstraße 12, 04109, Leipzig, Germany

    Jan-Niklas Meier & Paul Lehmann

  2. Institute for Empirical Economic Research, Leipzig University, Grimmaische Straße 12, 04109, Leipzig, Germany

    Bernd Süssmuth

  3. Leipzig University, Leipzig, Germany

    Stephan Wedekind

Authors
  1. Jan-Niklas Meier
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Paul Lehmann
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Bernd Süssmuth
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Stephan Wedekind
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Jan-Niklas Meier.

Ethics declarations

Conflict of interest

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendices

1.1 Appendix A: Summary statistics

See Tables 6, 7

Table 6 Panel data summary statistics
Full size table
Table 7 Correlation matrix
Full size table

1.2 Appendix B: Extended model specifications

See Tables 8, 9

Table 8 State-level policies
Full size table
Table 9 Regional policies and zoning details
Full size table

1.3 Appendix C: Robustness checks

See Tables 10, 11

Table 10 Subsample of rural counties
Full size table
Table 11 Conditional fixed-effects poisson regression
Full size table

1.4 Appendix D: Time delay of policy impacts

See Figure 10.

Fig. 10
figure 10

Impact response analysis

Full size image

1.5 Appendix E: Marginal effects

1.5.1 Specification: “state interactions”, Table 8 column 2

See Table 12, Figure 11.

Table 12 Average marginal effects (AME) of State Policies
Full size table
Fig. 11
figure 11

Marginal effects conditional on geographical characteristics

Full size image

1.5.2 Specification: “regional interactions (A)”, Table 9 column 1

See Table 13, Figures 12, 13

Table 13 Marginal effects (ME) of total designated area conditional on type of planning
Full size table
Fig. 12
figure 12

Marginal effects conditional on exclusive planning

Full size image
Fig. 13
figure 13

Marginal effect of exclusive planning

Full size image

1.5.3 Specification: “regional interactions (B)”, Table 9 column 2

See Table 14, Figure 14.

Table 14 Marginal effects (ME) of total designated area under exclusive planning
Full size table
Fig. 14
figure 14

Marginal effects conditional on avg. capacity density and exclusive planning

Full size image

1.5.4 Main specification: RE model, Table 3 column 3

See Figure 15.

Fig. 15
figure 15

Conditional marginal effects of wind power potential (based on RE estimates)

Full size image

1.5.5 Marginal effects for Northern and Southern German Counties

See figure 16

Fig. 16
figure 16

Marginal effects conditional on geographic region

Full size image

1.6 Appendix F: Micrositing

See Figures 17, 18

Fig. 17
figure 17

Example for micrositing in areas of different shape

Full size image
Fig. 18
figure 18

Real world example for micrositing in areas of different shape

Full size image

1.7 Appendix G: Capacity density

See Figure 19.

Fig. 19
figure 19

Development of the average capacity density per state

Full size image

1.8 Appendix H: Calculation of the county-specific and time-variant RES support level

See Table 15.

Table 15 Input values and calculation steps
Full size table

1.9 Appendix I: State-level policies

See Table 16, 17, 18.

Table 16 Forest Ban
Full size table
Table 17 Setback rule
Full size table
Table 18 Expansion goal
Full size table

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Meier, JN., Lehmann, P., Süssmuth, B. et al. Wind power deployment and the impact of spatial planning policies. Environ Resource Econ 87, 491–550 (2024). https://doi.org/10.1007/s10640-023-00820-3

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10640-023-00820-3

Keywords

JEL Classification

Search

Navigation