Subsidies for Renewable Energies in the Presence of Learning Effects and Market Power

Transcription

1 Subsidies for Renewable Energies in the Presence of Learning Effects and Market Power Department of Economics,, Germany Centre for Energy and Environmental Markets, UNSW Public Seminar

2 Motivation

3 Motivation Promotion of renewable energy sources has moved into the center of attention of many OECD economies energy policy, driven by concerns over the security of energy supply, global climate change, etc.

4 Motivation Promotion of renewable energy sources has moved into the center of attention of many OECD economies energy policy, driven by concerns over the security of energy supply, global climate change, etc. Instruments: quotas combined with tradable green certificates (UK, Italy, Australia), tenders (Ireland) and feed-in tariffs (Germany, Spain, Denmark)

5 Motivation Promotion of renewable energy sources has moved into the center of attention of many OECD economies energy policy, driven by concerns over the security of energy supply, global climate change, etc. Instruments: quotas combined with tradable green certificates (UK, Italy, Australia), tenders (Ireland) and feed-in tariffs (Germany, Spain, Denmark) In Europe, feed-in tariffs particularly effective in promoting the rapid expansion of RES-E capacity and production

6 Motivation Promotion of renewable energy sources has moved into the center of attention of many OECD economies energy policy, driven by concerns over the security of energy supply, global climate change, etc. Instruments: quotas combined with tradable green certificates (UK, Italy, Australia), tenders (Ireland) and feed-in tariffs (Germany, Spain, Denmark) In Europe, feed-in tariffs particularly effective in promoting the rapid expansion of RES-E capacity and production Claim: policy intervention is justified in the early stage of RES-E use to spur learning by doing and enable RES-E producers to move downwards on their learning curves, until they become competitive wrt conventional electricity producers

7 Motivation

8 Motivation Figure: Development of electricity generation from renewable electricity in the EU-27 (excluding hydropower)

9 Contributions

10 Contributions We study first-best and second-best policies, taking account of three important features of European electricity markets:

11 Contributions We study first-best and second-best policies, taking account of three important features of European electricity markets: 1 Oligopolistic competition in the fossil fuel electricity sector 2 Learning by doing in the RES-E equipment industry 3 Oligopolistic competition in the RES-E equipment industry

12 Division of the electricity network in Germany

13 Division of the electricity network in Germany

14 Learning curves for wind turbines and PV modules

15 Learning curves for wind turbines and PV modules Source: Grübler et al., 1999

16 Market shares in the wind turbine industry

17 Market shares in the wind turbine industry Six market leaders in the wind turbine industry: Vestas (Denmark), GE Wind (US), Gamesa (Spain), Enercon (Germany), Suzlon (India), Siemens (Germany) 85% of world market in 2008 Smaller expanding players: Sinovel (China), Acciona (Spain), Goldwind (China), Nordex (Germany) However, many turbine manufacturers are still mainly active in their domestic and neighboring markets e.g. Enercon, Vestas, and Siemens supply over 50% of the German, Dutch, and UK markets, respectively Source: BTM-C, 2009

18 The firms: fossil fuel utilities

19 The firms: fossil fuel utilities Cost K t (k t ) K t (k t ) > 0, K t (k t ) > 0

20 The firms: fossil fuel utilities Cost K t (k t ) K t (k t ) > 0, K t (k t ) > 0 Profit πt F (k t ) = P t (Q t )k t K t (k t ) τ t k t P t (Q t ) : downwards sloping inverse demand function for electricity Q t : total electricity production τ t : emission tax t = 1, 2

21 The firms: RES-E generators

22 The firms: RES-E generators Cost C t (q t, x) with x : location parameter Cq t > 0, C t x > 0, C qq t > 0, Cq x t > 0, C t x x > 0

23 The firms: RES-E generators Cost C t (q t, x) with x : location parameter Cq t > 0, C t x > 0, C qq t > 0, Cq x t > 0, C t x x > 0 Profit π G t (q t, x) = P t (Q t )q t C t (q t, x) b t b t : RES-E equipment price Q t = X t 0 q t ( x)d x + mk t P t (Q t )q t (X t ) C t (q t, X t ) b t = 0 : zero-profit condition X t : marginal RES-E producer

24 The firms: RES-E equipment producers

25 The firms: RES-E equipment producers Cost Γ 1 (y 1 ), Γ 2 (y 2, L) with L = y 1 + (n 1)εỹ 1 Γ t y t > 0, Γ t y ty t > 0,Γ 2 L < 0, Γ2 y 2L < 0, Γ2 LL > 0

26 The firms: RES-E equipment producers Cost Γ 1 (y 1 ), Γ 2 (y 2, L) with L = y 1 + (n 1)εỹ 1 Γ t y t > 0, Γ t y ty t > 0,Γ 2 L < 0, Γ2 y 2L < 0, Γ2 LL > 0 Profit π E t (y 1, y 2 ) = [B 1 (X 1 )+σ 1 ]y 1 Γ 1 (y 1 )+δ [ [B 2 (X 2 )+σ 2 ]y 2 Γ 2 (y 2, L) ] B t (X t ) : inverse demand function for RES-E equipment defined by zero-profit condition of marginal RES-E generator X t = ny t : total number of RES-E equipment with n firms in the RES-E equipment industry σ t : output subsidy δ : discount factor

27 Welfare

28 Welfare Q1 W = 0 X1 P 1 (Q)dQ mk 1 (k 1 ) nγ 1 (y 1 ) D 1 (mk 1 ) + δ X2 0 [ Q2 0 0 C 1 (q 1, x)d x P 2 (Q)dQ mk 2 (k 2 ) ] C 2 (q 2, x)d x nγ 2 (y 2, L) D 2 (mk 2 ) Emission damage: D t (mk t ) with D t(mk t ) > 0, D t (mk t ) 0 Number of fossil fuel utilities: m

29 FOC for profit maximization

30 FOC for profit maximization Fossil fuel firms: P t (Q t ) + P t(q t )k t K t (k t ) τ t = 0 (1) RES-E generators: RES-E equipment producers: P t (Q t ) C t q(q t, x) = 0 x [0, X t ] (2) B 1 (X 1 ) + B 1(X 1 )y 1 + σ 1 Γ 1 y 1 (y 1 ) (3) [ +δ B 2(X 2 )(n 1) ỹ ] 2 Γ 2 y L(y 2, L) = 0 1 B 2 (X 2 ) + B 2(X 2 )y 2 + σ 2 Γ 2 y 2 (y 2 ; L) = 0 (4)

31 FOC for welfare maximization

32 FOC for welfare maximization Welfare: W kt = P t (Q t ) K t (k t ) D t(mk t ) = 0 (5) W qt = P t (Q t ) Cq(q t t, x) = 0 x [0, X t ] (6) W y1 = P 1 (Q 1 )q 1 (X 1 ) C 1 (q 1, X 1 ) Γ 1 y 1 (y 1 ) (7) δ [ Γ 2 L(y 2, L)(1 + (n 1)ε) ] = 0 W y2 = P 2 (Q 2 )q 2 (X 2 ) C 2 (q 2, X 2 ) Γ 2 y 2 (y 2, L) = 0 (8)

33 Optimal Policy

34 Optimal Policy Optimal emission tax in both periods + {}}{{}}{ τt = P t(q t ) kt + D t(mk t )

35 Optimal Policy Optimal emission tax in both periods + {}}{{}}{ τt = P t(q t ) kt + D t(mk t ) Optimal subsidy in period 1 + {}}{{}}{{}}{ σ1 = B 1(X 1 )y1 δ(n 1)εΓ 2 L(y2, L ) δb 2(X 2 )(n 1) ỹ 2 y 1

36 Optimal Policy Optimal emission tax in both periods + {}}{{}}{ τt = P t(q t ) kt + D t(mk t ) Optimal subsidy in period 1 + {}}{{}}{{}}{ σ1 = B 1(X 1 )y1 δ(n 1)εΓ 2 L(y2, L ) δb 2(X 2 )(n 1) ỹ 2 y 1 Optimal subsidy in period 2 {}}{ σ2 = B 2(X 2 )y2

37 Optimal Policy

38 Optimal Policy Optimal tax in both periods corrects for marginal damage caused by pollution and the too low level of output due to oligopolistic competition in the fossil-fuel industry Optimal subsidy in the first period corrects for the output contraction due to oligopolistic competition, the strategic output expansion of the firms in the first period in order to shift their reaction curves outwards in the second period, and the learning spill-overs neglected by individual firms Optimal subsidy in the second period only corrects for the output contraction due to oligopolistic competition the the RES-E equipment industry

39 The impact of market structure on the policy instruments

40 The impact of market structure on the policy instruments The impact of market structure in the fossil fuel industry τ 1 m > 0 τ 2 m > 0 σ 1 m 0 σ 2 m 0

41 The impact of market structure on the policy instruments The impact of market structure in the fossil fuel industry τ 1 m > 0 τ 2 m > 0 σ 1 m 0 σ 2 m 0 The impact of market structure in the RES-E equipment industry τ 1 n < 0 τ 2 n < 0 σ 1 n < 0 σ 2 n < 0

42 Feed-In Tariffs

43 Feed-In Tariffs RES-E generators receive a feed-in tariff ζ t per unit of electricity produced in each period Feed-in tariffs paid by the government Exogenous emission tax No subsidy in the RES-E equipment sector Profit of RES-E generators π G t (q t, x, ζ t ) = ζ t q t C t (q t, x) b t

44 Second-Best Optimal Policy Feed-in tariff depends on five terms:

45 Second-Best Optimal Policy Feed-in tariff depends on five terms: 1 P t(q t): electricity price in t = 1, 2

46 Second-Best Optimal Policy Feed-in tariff depends on five terms: 1 P t(q t): electricity price in t = 1, 2 2 [ +/ {}}{{}}{ D t(mk t) τ t + P t (Q t)k t]: tax rate, marginal damage and the degree of market power in the fossil fuel industry in t = 1, 2

47 Second-Best Optimal Policy Feed-in tariff depends on five terms: 1 P t(q t): electricity price in t = 1, 2 2 [ 3 +/ {}}{{}}{ D t(mk t) τ t + P t (Q t)k t]: tax rate, marginal damage and the degree of market power in the fossil fuel industry in t = 1, 2 + {}}{ δγ 2 L(n 1)ε: learning by doing and the degree of learning spill-overs in the RES-E equipment industry

48 Second-Best Optimal Policy Feed-in tariff depends on five terms: 1 P t(q t): electricity price in t = 1, 2 2 [ 3 4 +/ {}}{{}}{ D t(mk t) τ t + P t (Q t)k t]: tax rate, marginal damage and the degree of market power in the fossil fuel industry in t = 1, 2 + {}}{ δγ 2 L(n 1)ε: learning by doing and the degree of learning spill-overs in the RES-E equipment industry {}}{ BX 1 1 (X 1, ζ 1)y 1 + equipment industry in t=1 + {}}{ δb 2 X 2 (X 2, ζ 2)(n 1) ỹ2 y 1 : strategic effects in the RES-E

49 Second-Best Optimal Policy Feed-in tariff depends on five terms: 1 P t(q t): electricity price in t = 1, 2 2 [ / {}}{{}}{ D t(mk t) τ t + P t (Q t)k t]: tax rate, marginal damage and the degree of market power in the fossil fuel industry in t = 1, 2 + {}}{ δγ 2 L(n 1)ε: learning by doing and the degree of learning spill-overs in the RES-E equipment industry {}}{ BX 1 1 (X 1, ζ 1)y 1 + equipment industry in t=1 + {}}{ δb 2 X 2 (X 2, ζ 2)(n 1) ỹ2 y 1 : strategic effects in the RES-E {}}{ B 2 X 2 (X 2, ζ 2)y 2: strategic effects in the RES-E equipment industry in t=2

50 Welfare implications

51 Welfare implications Table: Welfare loss of a second-best feed-in tariff policy versus the first-best policy Welfare Exog. tax Welfare Welfare loss (fb) rate (sb) (%) Oligopoly in the fossil-fuel τ industry only τ Oligopoly in the fossil-fuel and RES-E τ equipment industries τ Functional forms: P(Q G t, kt ) = A B(mkt + QG t ), C t (q t ) = 1 2 c(qt + f x)2, K t (k t ) = h 2 k2 t, Γ1 (y 1 ) = γ 2 y 2 1, Γ 2 (y 2, L) = γ 4 (y 2 bl) 2 + γ 4 y 2 2, Dt (mkt ) = d 2 mk2 t Parameter values (baseline): A = 10, B = 0.5, h = 0.1, b = 0.1, γ = 0.2, c = 0.5, f = 0.5, ε = 0.5, d = 1

52 The impact of market structure on the feed-in tariffs

53 The impact of market structure on the feed-in tariffs Increasing the number of firms in the fossil fuel industry Ζ1,Ζ

54 The impact of market structure on the feed-in tariffs Increasing the number of firms in the fossil fuel industry Ζ1,Ζ Increasing the number of firms in the RES-E equipment industry Ζ1,Ζ

55 Conclusions and policy recommendations

56 Conclusions and policy recommendations FITs for renewable electricity generators may be justified in the presence of market power and learning spill-overs, if first-best optimal policies are not available to the regulator.

57 Conclusions and policy recommendations FITs for renewable electricity generators may be justified in the presence of market power and learning spill-overs, if first-best optimal policies are not available to the regulator. The welfare loss of second best FITs wrt a first best policy is considerably higher when there is imperfect competition in the RES-E equipment industry, since FITs are not very effective in internalizing pollution damage and the strategic effects in the RES-E equipment and the fossil-fuel industry.

58 Conclusions and policy recommendations FITs for renewable electricity generators may be justified in the presence of market power and learning spill-overs, if first-best optimal policies are not available to the regulator. The welfare loss of second best FITs wrt a first best policy is considerably higher when there is imperfect competition in the RES-E equipment industry, since FITs are not very effective in internalizing pollution damage and the strategic effects in the RES-E equipment and the fossil-fuel industry. FITs should be increased in response to increasing competition in the fossil fuel industry and decreased in response to increasing competition in the RES-E equipment sector.

59 Conclusions and policy recommendations FITs for renewable electricity generators may be justified in the presence of market power and learning spill-overs, if first-best optimal policies are not available to the regulator. The welfare loss of second best FITs wrt a first best policy is considerably higher when there is imperfect competition in the RES-E equipment industry, since FITs are not very effective in internalizing pollution damage and the strategic effects in the RES-E equipment and the fossil-fuel industry. FITs should be increased in response to increasing competition in the fossil fuel industry and decreased in response to increasing competition in the RES-E equipment sector. With imperfect competition in the RES-E equipment industry, FITs should be higher in the second than in the first period.

60 References Emission taxation in imperfectly competitive markets cf. Buchanan, 1969, Lee, 1975, Barnett, 1980 Learning-by-doing and learning spill-overs in imperfectly competitive markets cf. Spence, 1981, Fudenberg and Tirole, 1983, Goulder and Mathai, 2000, Bramoullé and Olson, 2005, Fischer and Newell, 2008 Learning effects in the renewable energy sector cf. Grübler et al., 1999, Hansen et al., 2003, Junginger et al., 2005, Isoard and Soria, 2001, McDonald and Schrattenholzer, 2001, van der Zwan and Rabl, 2004, Neij, 1997 and 1999

61 Composition of ỹ 2 y 1 ỹ 2 y 1 = Γ 2 y 2 Γ 2 y 2 L (B 2 + B 2 ỹ2) + ε[γ 2 2 y 2 LΓ y 2 y Γ 2 2 y 2 L(2B 2 + B 2 y 2)] y Γ 2 2 y 2 y (n + 1)(B 2 2 )2 + Γ 2 y 2 y (2B B 2 y 2) + Γ 2 y 2 y (nb (n 1)B 2 ỹ2) (y 2 + (n 1)ỹ 2 )B 2 B 2 ỹ 2 y 1 : Comparative statics effect of increasing output of firm 1 in the first period on output of the other firms in the second period ỹ 2: Output of all other turbine firms in the second period Assuming that the spill-over coefficient ε is not too large, by convexity of the inverse demand function this expression becomes negative. This implies that an increase in output by firm 1 in the first period has a negative effect on the output decisions of the other firms in the second period

62 Second-best optimal feed-in tariffs ζ oc 1 =P 1 (Q 1 ) + [D 1 (mk 1) τ 1 + P H 2 m k 1 H 1 (Q ζ 1 m k 1 1)k 1 ] 1 ζ 2 C 2 H 1 C 1 H 2 [B 1 X (X 1 1, ζ 1 )y 1 + δb 2 X (X 2 2, ζ 2 )(n 1) ỹ y 2 + δγ 2 H 1 n 1 H L y (n 1)ɛ] ζ 2 n y 1 2 ζ 1 1 C 2 H 1 C 1 H 2 + δ[d 2 (mk 2) τ 2 + P H 2 m k 2 H 2 (Q ζ 1 m k 2 2)k 2 ] 1 ζ 2 C 2 H 1 C 1 H 2 δb 2 X 2 (X 2, ζ 2 )y 2 H 1 n y 2 ζ 2 H 2 n y 2 ζ 1 C 2 H 1 C 1 H 2 (9) where C 1, C 2, H 1, and H 2 denote the reaction of green electricity production when the subsidy rate changes in a particular period, i.e. C 1 = q 1 (X 1 ) X 1 + X 1 q 1 ( x) ζ 1 0 d x, C ζ 2 = q 1 (X 1 ) X 1, 1 ζ 2 H 1 = q 2 (X 2 ) X 2 and H ζ 2 = q 2 (X 2 ) X 2 + X 2 1 ζ 2 0 q 2 ( x) d x. ζ 2

63 Numerical example: functional forms Table: Functional forms Functional form Description C t (q t ) = 2 1 c(qt + f x)2 Cost function of the RES-E generators K t (k t ) = h 2 k2 t Cost function of the fossil fuel firms Γ 1 (y 1 ) = γ 2 y 1 2 Cost function of the RES-E equipment producers in t=1 Γ 2 (y 2, L) = γ 4 (y 2 bl) 2 + γ 4 y 2 2 Cost function of the RES-E equipment producers in t=2 D t (mk t ) = d 2 mk2 t Pollution damage P t (Qt G, kt ) = A B(mkt + QG t ) Demand function for electricity

64 The impact of market structure in the fossil fuel industry Ζ1,Ζ2 p1,p2 Q1,Q mk1,mk2 G1,G b1,b X1,X Figure: Impact Johanna ofreichenbach, increasing Tillthe Requate number University of firms of Kiel in the fossil fuel

65 Figure: Impact Johanna ofreichenbach, increasing Tillthe Requate number University of firms of Kiel in the RES-E The impact of market structure in the RES-E equipment industry Ζ1,Ζ2 p1,p2 Q1,Q mk1,mk G1,G b1,b X1,X