Wind electricity subsidies — A windfall for landowners? Evidence from a feed-in tariff in Germany

Highlights

• We investigate the incidence of subsidies for wind energy on agricultural land prices due to a feed-in tariff in Germany.

• For identification we exploit variation in wind strength across 260 non-urban counties in combination with the introduction of the subsidy.

• We find that land prices increased by roughly 1,100 euro in high-wind areas after the introduction of the subsidy.

• About 18% of expected wind turbine profits are capitalized into land prices.

• Based on the estimated incidence, we calculate that wind turbine subsidies account for 4% of agricultural income in 2007.

Abstract

Subsidies for renewable energy sources are increasing around the globe and amounted to more than 100 billion euro in 2013. This study aims to answer whether the subsidies only ensure that green electricity plants are profitable or whether other market participant – as, for example, landowners – benefit from the subsidy in the form of windfall gains as well. To identify the causal effect of the subsidies, we investigate the impact of the introduction of a price guarantee in the form of a feed-in tariff for wind electricity in Germany on land prices. We employ two different approaches. Both approaches exploit quasi-experimental variation in wind strength across 260 non-urban counties in combination with the introduction of the subsidies. Based on a difference-in-differences design, we find that land prices increased by roughly 1100 euro in high-wind areas after the introduction of the subsidy. Further, using an instrumental variable estimator we document that around 18% of expected wind turbine profits are capitalized into land prices. Finally, we show that wind turbine subsidies account for 4% of overall agricultural income in 2007.

Introduction

Most countries subsidize investment in renewable energy sources (RES) as a higher share of green electricity is seen as key for alleviating global warming. Taking into account all the various support schemes, the International Energy Agency estimates that in 2013 around 100 billion euro (121 billion US dollars) were spent worldwide to promote green energy (International Energy Agency, 2014). These enormous amounts raise concerns about the distributional consequences and the efficiency of the subsidies. In particular, it is crucial to understand who benefits from the subsidies. Is it the investor, and consequently does the subsidy ensure that green electricity plants are profitable? Or to what extent do other market participants – as, for example, landowners – benefit from the subsidy in the form of windfall gains?¹ This paper aims to provide empirical evidence on this incidence question by investigating the effect of onshore wind turbine subsidies in Germany on administrative transaction prices for agricultural land between 1997 and 2004.

The setting in Germany is particularly suitable for our analysis. First, RES subsidies are important. In 2012, about 12 billion euro were spent on these subsidies, roughly the equivalent of 20% of Germany's tax revenue on corporate profits. Second, Germany uses a price guarantee for green electricity in the form of a feed-in tariff to foster investment in RES. The feed-in tariff was introduced in 2000 with the adoption of the Renewable Energy Act (REA). The central mechanism of a feed-in tariff is a guaranteed, fixed wholesale price for green electricity for a specific time period. To date, this support scheme is the most commonly used policy measure around the world (REN, 2017). Third, descriptive evidence based on aggregate statistics suggests that the REA had a sizable impact on the energy market. Following the introduction of the REA, electricity produced by RES, as a share of overall electricity consumption, increased from 6.2% in 2000 to 23.7% in 2012. Over the entire period, at least half of the overall electricity generated by RES came from onshore wind turbines.² Finally, landowners and investors are likely to be the main beneficiaries as there is little labor involved in the electricity generated by wind turbines.³

In order to identify the causal effect of the subsidy on land prices, we propose two different approaches which exploit quasi-experimental variation in wind strength in combination with the introduction of the REA in 2000. First, we apply a difference-in-differences (DiD) estimator by comparing land prices in high-wind and low-wind non-urban counties before and after the introduction of wind electricity subsidies. Since this design only identifies the causal effect of the REA on land prices but does not allow the quantification of the incidence of the subsidy, we estimate in a second step the effect of expected wind turbines profits on land prices. This specification is motivated by Titman's work on the price of vacant land under uncertainty (Titman, 1985) which suggests that the price of each field on which a wind turbine can potentially be built increases with the subsidy. As expected wind turbines profits are not observed, we construct the profits based on simulated expected profits of a median-technology wind turbine for each county and year. We address the potential measurement error in our simulation due to the assumed technological choice, which would bias the OLS estimator, by employing an instrumental variable (IV) strategy. The excluded instrument is constructed using variation in wind strength across counties in combination with the introduction of the REA. To provide a deeper understanding of the incidence result, we analyze in the final part of the paper how large the additional income generated by wind turbines subsidies is in comparison to overall agricultural income.

Our results suggest positive incidence effects. In the DiD estimation, we find that land prices per hectare increased by around 1300 euro for each additional meter per second (m/s) wind strength in a county after the introduction of the REA. In the IV estimation we show that around 18% of expected wind turbine profits are capitalized into land prices. The results are robust across a wide range of specifications. The estimated incidence share translates – based on 2004 values – into capitalized wind turbine subsidies of 4000 euro per hectare or 25% of the average land price per hectare. Finally, using the estimated incidence share and taking into account the characteristics of the German agricultural land market, we show that wind turbine subsidies increased agricultural income by 4% on average.

Our paper contributes to previous literature in several ways. First, we add to the literature on the incidence of subsidies in agricultural land prices. Our estimated incidence share is at the lower end of estimates for agricultural subsidies.⁴ For the US, Kirwan (2009) estimates an incidence share of 25%. Hendricks et al. (2012) suggest that the long-run incidence in the US may be up to 40% as inertia in farmland rental rates and different types of tenancy agreements are likely to have biased prior estimates. This is similar to findings by Roberts et al. (2003).⁵ Ciaian and Kancs (2012) report that in OECD countries around 20% of agricultural subsidies are reaped by landowners. Most comparable to our study is the work by Breustedt and Habermann (2011). They estimate an incidence share of 38% for agricultural subsidies for the German state of Lower Saxony.

Second, we build on the literature that quantifies the distributional impact of environmental policies. Fullerton (2011) discusses six ways in which environmental policies (mainly taxation) may have distributional impact, which are all likely to be regressive. Empirical literature confirms this presumption for environmental taxes Parry, 2004, Metcalf, 1999; Hassett et al., 2009, Grainger and Kolstad, 2010. The driving force behind the distributional impact of a carbon tax is the fact that low income households spend a large share of their budget on polluting goods (Grainger and Kolstad, 2010). Metcalf (1999) suggests targeted tax cuts to make the policy distributionally neutral. Parry et al. (2005) question the regressive impact of environmental policies by pointing out that poorer households bear a disproportionate share of environmental risk, which is reduced by these policies. However, the authors suggest also that environmental policies are capitalized into housing prices, which benefit mainly richer households as they are more likely to own their home. The distributional impact of subsidies for renewable energies has received less attention, although they are commonly used in practice and recent theoretical literature has provided a rationale for the observed policies. Eichner and Runkel (2014), for example, show that subsidizing RES in addition to taxing pollution reduces the distortion of the tax-subsidy system.⁶ One of the few empirical studies that investigates subsidies for renewable energies is Groesche and Schroeder (2014). They focus on subsidized photovoltaic plants on owner-occupied houses in Germany and find that the German REA is mildly regressive.

Finally, our results add to literature that discusses the efficiency of different policy instruments to promote renewable energies (e.g. Menanteau et al., 2003, Haas et al., 2011, Requate, 2014. Whether a price-based (such as a feed-in tariff) or a quantity-based system (such as tradable green electricity certificates and quotas) is preferable to foster investment in RES is often evaluated through three criteria: costs, installed capacity, and technological development (see Menanteau et al., 2003). Different distributional implications are not discussed, although they are important and likely to differ.

The outline of the paper is as follows. Section 2 describes the relevant institutions and in particular the REA. The data is presented in Section 3, followed by the methodology in Section 4. Results on capitalization and incidence share are presented in Section 5, while the distributional consequences of the wind turbine subsidies are discussed in Section 6. Section 7 concludes.