Political Economy of Biofuels

Abstract

While timber and other biomass have been the main sources of fuel for millennia, there has been an increasing emphasis on growing crops and converting feedstock to liquid fuels (Rajagopal et al. 2009) or for use in power plants. These new fuels were induced by government policies and often require a diversion of resources from agricultural to energy production. Analyzing the performance of biofuels and biofuel policies requires a political economic lens—this chapter will provide such a framework to assess biofuels.

Appendix

The framework can be generalized to establish policies internationally. We assume that the country has I interest groups and let i = 1, …, I an interest group indicator. The relevant interest groups in the context of biofuels include food consumers, food producers, fuel consumers, biofuel producers, fossil fuel producers, environmentalists, automobile companies, and macro policy makers. Each interest group is assumed to have a political weight, β _i , as well as welfare function W _i  = g _i (X) where X is a K dimensional vector of indicators of well-being (the index k = 1, …, K), and the element of the vector X _i is x _k . These indicators can be both quantities and prices, and they reflect performance relative to a benchmark. They include changes relative to a benchmark in the prices of food and fuel, GHGE, government revenue, balance of trade, and food security (measured by the amount imported from less secure regions), as well as other indicators. Thus, if k = 1 indicates food price, x ₁ is a change in the price of food relative to a benchmark, and g ₁(x ₁, …, x _k ) represents the effect of the changes in indicators that affect the well being of the first interest group.

Each indicator is a function of biofuel policy variables. Let Y be an N dimensional vector of policies n = 1, …, N, and the nth element of Y is y _n . Each policy indicator is a function of the biofuel policy variables, and let the function x _k  = φ _k (Y) denote this functional relationship. The policy variables we consider may include the level of biofuel tax, a biofuel mandate, biofuel subsidies, biofuel tariff, GHGE regulations related to biofuel, regulation on the use of biofuel (the blend wall). The relationship between the indicators and policy is determined by market interaction and other constraints that affect the economy, which are not specified here. To simplify our analysis, we assume that there is a direct functional relationship between welfare of each group and policy measures, denoted by f _i (Y) for i = 1, …, I. Thus, W _i  = g _i (x ₁(Y), …, x _k (Y)) = f _i (Y). With this notation, if i = 1 is food consumers, f ₁(Y) is the effect of a proposed policy y on the well being of consumers (through its effect on the different performance measures).

We will assume like a few of the models overviewed by Rausser et al. (2011) that the determination of a policy parameter can be determined by the maximization of a political economic objective function S = h(f ₁(Y), β ₁, …, f _I (Y), β _I ), which is assumed to be well-behaved. For example, this objective function can be:

• Linear $S={\displaystyle \sum}_{i=1}^I{\beta}_i{f}_i(Y)$

• Log linear $S={\displaystyle \sum}_{i=1}^I{\beta}_i \log {f}_i(Y)$

The determination of optimal policy is subject to a variety of constraints, including budgetary limits, biophysical limitations on a feasible range of policy parameters, etc. We assume that there are M constraints and the indicator of constraints is m = 1, …, M, and the mth constraint is c _m (y ₁, …, y _n ) = 0 where c _m is a well-behaved function of the policy parameters.

Thus, the political system will result in policy parameters y _n that solve:

$$\underset{y_1,\dots {y}_n}{ \max }h\left({f}_1(Y),{\beta}_1,\dots, {f}_I(Y),{\beta}_I\right) \mathrm{s}.\mathrm{t}. {c}_m\left({y}_1,\dots, {y}_n\right)=0 \mathrm{where} m=1,\dots, M $$

The first-order condition that will determine the policy parameter y _i is:

$$ {\displaystyle \sum_{i=1}^I\frac{\partial h}{\partial {f}_i(Y)}}\frac{\partial {f}_i(Y)}{\partial {y}_n}-{\displaystyle \sum_{m=1}^M{\lambda}_m\frac{\partial {c}_m}{\partial {y}_n}}=0 $$

(11.1)

Where n = 1,…N where λ _m is the shadow price of the mth constraint. The first-order condition (1) suggests that the optimal level of y _n is where the weighted sum of marginal contributions is equal to the sum of the marginal cost it imposes on the constraints. In the linear scenario, Eq. (11.1) becomes (11.2).

$$ \displaystyle \sum_{i=1}^I {\beta_{i}}\frac{\partial {f}_i(Y)}{\partial {y}_n}-{\displaystyle \sum_{m=1}^M{\lambda}_m\frac{\partial {c}_m}{\partial {y}_n}}=0 $$

(11.2)

Equation (11.2) suggests that policies that are desired by interest groups with large political weight will gain more support. However, political weight is not the only factor that affects policy determination. It is affected by the impact on marginal benefit to various groups and to what extent the policy is constrained by the factors mentioned above.