Energy efficiency is increasingly being recognized as a resource warranting aggressive public investment. Policy makers, regulators, and utilities are currently taking a serious look at policy mechanisms that are being considered or are already in place in many states to encourage utilities to pursue energy efficiency. The State of California committed an unprecedented sum of $2.2 billion in ratepayer funds to utility-based energy efficiency programs from 2006 through 2008; the state finalized in 2007 the shared-savings incentive mechanism for the 2006-2008 programs and beyond.

This dissertation explores (i) the implementation of utility-based energy efficiency programs that employ a shared-savings incentive mechanism and (ii) the interest-group politics of shaping the incentive mechanism. California energy efficiency programs and the associated incentive rulemaking process are examples of both of these.

The first study presents an economic model for the implementation of the programs, in which a regulator adopts an energy savings target and a shared-savings incentive mechanism before a utility firm proposes program funding, gets it authorized, and begins to manage it. The study reveals that (i) each utility firm requires a certain minimum incentive rate to ensure that the firm will be encouraged to achieve the energy savings target, eventually bringing non-negative net earnings to the firm and non-negative bill savings to its customers; (ii) what distinguishes the asymmetric from the symmetric information case is that a utility firm with virtually any program management efficiency is now left with strictly positive net earnings; (iii) depending on market and regulatory circumstances, a higher-than-minimum incentive rate can be warranted to achieve not only a greater net social benefit but also greater bill savings for customers; and (iv) a multiple-tier marginal incentive rate, if appropriately tiered, can generate higher energy savings performance than the single-tier counterpart with the same program funding could have and can achieve a Pareto-superior allocation.

The second study proposes a positive economic model with which to describe a deliberative policy making process that requires groups with conflicting interests to communicate with each other and to propose their preferred policy implementation details before policy makers determine a policy, taking these proposals into account. The model assumes that each group balances a greater rent derivable from making a more advantageous proposal with a greater political effort it will have to make to eliminate its rival's greater political resistance. This political resistance increases both with the rival's implied utility loss associated with the proposal and with the rival's political effectiveness. The model also assumes that the policy makers establish a policy on the basis of their judgment about the groups' proposals and that the bias in the judgment is known to the interest groups a priori. The study predicts that an increase in the political effectiveness of either interest group will make both of them less polarized in their proposals, reducing political costs incurred as a result of the political competition. It also predicts that a consensus over a policy can be reached a priori by the two groups if both of them are sufficiently effective in politics.

Analyses of the economic models developed in this dissertation offer two major policy implications for California energy efficiency programs. First, a higher-than-adopted incentive rate would achieve not only a greater net social benefit but also greater bill savings for customers. Second, social efficiency would be better achieved by customizing incentive mechanisms for individual utilities and updating them on a regular basis.