Accessible Unlicensed Requires Authentication Published by De Gruyter January 30, 2019

Refunding Emissions Taxes: The Case For A Three-Part Policy

Philippe Bontems


This paper examines theoretically whether by combining both output-based refunding and abatement expenditure-based refunding, it is possible to limit the negative consequences that a pollution tax implies for a polluting industry. We show that this is indeed the case by using a three-part policy where emissions are subject to a fee and where output and abatement expenditures are subsidized. When the industry is homogenous, it is possible to replicate the standard emission tax outcome by inducing a polluting firm to choose the production and emission levels obtained under any emission tax, without departing from budget balance. By construction, any polluter earns strictly more than under the standard tax alone without rebate, making this proposal more acceptable to the industry. When firms are heterogeneous, the refunding policy needed to replicate the standard emission tax outcome is personalized in the sense that at least the output subsidy should be type dependent and it is strictly preferred only from the industry’s point of view to a standard environmental tax. We also explore the implications of uniform three-part refunding policies for a heterogeneous industry.

JEL Classification: Q52; Q58; H23


A Comparative statics

Dropping the firm’s index for the sake of clarity and introducing a positive scale parameter θ for abatement cost function, the system eq. (1) rewrites as:


Differentiating totally this system and dropping arguments, we obtain:


where Δ=aeec+θ(aeeaqqaqe2)>0 under quasi-convexity of total cost c(.)+a(.,.). We obtain that:

qp=1Δaee>0and et=c+θaqqΔ<0.

Also, under assumption Assumption 1, we obtain that:

qt=1Δaqe<0and ep=aqeΔ>0.

Finally, assumption Assumption 2 allows to state that:

qθ=(aqaeeaeaqe)Δ<0and eθ=(aeaqqaqaqe))Δ>0.

B Proof of Proposition Proposition 2

To get qˆ(τ,s,f)=q and eˆ(τ,s,f)=e, in view of the FOCs eqs. (2) and (5), it is sufficient to take


Also, the subsidy s is given by the budget constraint eq. (3) that now writes:


or equivalently with R=(1δ)tne


where σ=1εa,eεa,q with εa,e=aee/a<0 and εa,q=aqq/a>0 and recalling that t=ae from eq. (1). Observe that σ=0 needs to be ruled out otherwise the budget equality constraint cannot hold.

Also, to preserve quasi-concavity of the polluter’s program, we need s < 1 which amounts to assume that


whenever is is positive.

Finally, by construction, the difference in terms of net profit is


by using the budget constraint.

C Proof of Proposition Proposition 5

To get qˆi=q and eˆi=e, in view of the FOCs eqs. (1) and (7), it is sufficient to take


Moreover, the budget constraint writes:


Using eq. (12) and replacing in eq. (13), we get


which implies


with σi=1εai,eεai,q.

Moreover, the difference between net profits is:


Replacing with the values obtained for the instruments τi and fi and recalling that t=aie, we obtain:


Summing over i and using eq. (14), we thus get ΠˆΠ=iσiaisi=δtE>0.

D Example

From the expression for individual production,


we get by summing over i:


Also from


we obtain similarly


Finally, for abatement expenditures we obtain:


The system to be solved in (τ,s,f) rewrites as follows:


Equation (16) can be expressed as


from which we deduce that


Also, equation (17) writes as


and replacing τ using eq. (19) allows to obtain


which simplifies into


where λkcσγ2σγ2+γˉ20. We can deduce f as a function of s:


When there is homogeneity, σγ2=0 and thus λ = 0 and we recover f=(1s)t as in Proposition 2.

Last, using eqs. (16) and (17), the budget constraint eq. (18) rewrites simply as


or using eq. (15) and the definition of E and Q:


This allows to compute s/(1s) as a function of f:


Hence, it follows that:


and replacing in eq. (20):


which amounts to solve a polynomial equation of degree 2 in f. We denote the two solutions f1 and f2. And the corresponding values of τ and s by, respectively, τ1, τ2, s1 and s2.

E Reallocation of production and pollution between firms

We compare qˆi and eˆi taken at the solution 1 (f1,τ1,s1) with their counterparts qi and ei in the standard tax benchmark. Let us start with quantities:


From eq. (19), we know that τ1=(f1t)γˉ and replacing in eq. (21), we get:


As solution 1 entails overtaxation of emissions (f1>t), we obtain that qˆi>qiγi<γˉ.

Now we compare the emission levels:


Using eq. (20) and replacing in eq. (22), we obtain:


Recall that λkcσγ2σγ2+γˉ2 so that by rearranging,


Once again, using that f1>t, it follows that eˆi>eiγi<γˉ+σγ2γˉ.


Aidt, T. S. 2010. “Green Taxes: Refunding Rules and Lobbying.” Journal of Environmental Economics and Management 60(1): 31–43.10.1016/j.jeem.2010.04.001Search in Google Scholar

Andreoni, J., and A. Levinson. 2001. “The Simple Analytics of the Environmental Kuznets Curve.”Journal of Public Economics 80(2): 269–86.10.1016/S0047-2727(00)00110-9Search in Google Scholar

Bonilla, J., J. Coria, K. Mohlin, and T. Sterner. 2015. “Refunded Emission Payments and Diffusion of NO x Abatement Technologies in Sweden.” Ecological Economics 116: 132–45.10.1016/j.ecolecon.2015.03.030Search in Google Scholar

Bontems, P., and J. M. Bourgeon. 2005. “Optimal Environmental Taxation and Enforcement Policy.” European Economic Review 49(2): 409–35.10.1016/S0014-2921(03)00036-9Search in Google Scholar

Cato, S. 2010. “Emission Taxes and Optimal Refunding Schemes with Endogenous Market Structure.” Environmental and Resource Economics 46(3): 275–80.10.1007/s10640-009-9340-2Search in Google Scholar

Conseil des Prélèvements Obligatoire. 2018. Les taxes affectées: des instruments à mieux encadrer. Report. .Search in Google Scholar

Coria, J., and K. Mohlin. 2017. “On Refunded Emissions Payment and Technological Diffusion.” Strategic Behavior and the Environment 6(3): 205–48.10.1561/102.00000068Search in Google Scholar

Fischer, C. 2001. Rebating Environmental Policy Revenues: Output-based Allocations and Tradable Performance Standards, 01–22. Washington, DC: Resources for the Future.Search in Google Scholar

Fredriksson, P. G., and T. Sterner. 2005. “The Political Economy of Refunded Emissions Payment Programs.” Economics Letters 87(1): 113–19.10.1016/j.econlet.2004.11.001Search in Google Scholar

Gersbach, H., and T. Requate. 2004. “Emission Taxes and Optimal Refunding Schemes.” Journal of Public Economics 88(3): 713–25.10.1016/S0047-2727(02)00215-3Search in Google Scholar

Hagem, C., M. Hoel, B. Holtsmark, and T. Sterner. 2015. Refunding Emissions Payments (No. dp-15-05).Search in Google Scholar

Macho-Stadler, I., and D. Perez-Castrillo. 2006. “Optimal Enforcement Policy and Firms’ Emissions and Compliance with Environmental Taxes.” Journal of Environmental Economics and Management 51(1): 110–31.10.1016/j.jeem.2005.06.001Search in Google Scholar

OECD. 2015. Towards Green Growth?: Tracking Progress. Paris: OECD Publishing. in Google Scholar

Sterner, T., and L. Höglund-Isaksson. 2006. “Refunded Emission Payments theory, Distribution of Costs, and Swedish Experience of NOx Abatement.” Ecological Economics 57(1): 93–106.10.1016/j.ecolecon.2005.03.008Search in Google Scholar

Sterner, T., and B. Turnheim. 2009. “Innovation and Diffusion of Environmental Technology: Industrial NOx Abatement in Sweden under Refunded Emission Payments.” Ecological Economics 68(12): 2996–3006.10.1016/j.ecolecon.2009.06.028Search in Google Scholar

Wang, H., and D. Wheeler. 2005. “Financial Incentives and Endogenous Enforcement in China’s Pollution Levy System.” Journal of Environmental Economics and Management 49: 174–96.10.1016/j.jeem.2004.02.004Search in Google Scholar

Published Online: 2019-01-30

© 2019 Walter de Gruyter GmbH, Berlin/Boston