Publication: RAND Journal of Economics
Publication Date: 22-MAR-07
Delivery: Immediate Online Access
Author: Ishii, Jun ; Yan, Jingming

Article Excerpt
An empirical model of the "make or buy" decision faced by independent power producers (IPPs) in restructured U.S. wholesale electricity markets is derived to analyze power plant investment decisions by major IPPs from 1996 to 2000. The estimated investment cost and expected profit functions are used to evaluate the effectiveness of divestiture programs (which sold utility power plants to IPPs) in encouraging greater IPP participation. The estimates and counterfuctual simulations indicate that a minimal amount of new plant investments were "crowded out" by divestiture and that divestiture encouraged greater (short-run) entry, especially among utility-affiliated IPPs.

1. Introduction

* Beginning with California in 1996, many state governments in the United States have enacted restructuring legislation aimed at transforming the electricity supply industry away from the traditional regulated structure toward a more competition-based marketplace. Historically, the industry has been dominated by vertically integrated investor-owned utilities (IOUs), whose regulated geographic monopolies controlled all three main sectors of the industry: generation, transmission and distribution (T&D) and retail services. One of the main goals of the restructuring process is to introduce competition into the electricity-generation sector and allow nonutility, independent power producers (IPPs) to invest and compete for market-based returns. (1) In order to facilitate the introduction of competition, many state policymakers have argued for and implemented a plan under which incumbent IOUs were enticed to sell their existing generation assets to entrant IPPs.

Policymakers felt that the divestiture of IOU generation assets would serve three purposes. First, it would help prevent the incumbent utilities from using their existing generation and T&D facilities to exert market power in the restructured generation sector. Second, divestiture would provide a market-based method of evaluating the "stranded cost" that IOUs would incur due to restructuring. (2) Many state restructuring programs required divestiture as a condition for state assistance in recovering stranded cost. Third, many policymakers believed that divestiture sales would help encourage greater entry and investment by IPPs. IPPs buying these divested assets would be able to participate immediately in these restructured markets. Moreover, divestiture could help encourage IPP participation by signaling a greater commitment to restructuring; a state that has transferred a greater amount of its existing generation supply out of the hands of regulated IOUs and into the hands of unregulated IPPs would be harder pressed putting the "genie back in the bottle."

The empirical evidence from the past 10 years of electricity restructuring in the United States has been mixed on the effectiveness of divestiture in achieving these goals. Much of the existing analysis of this evidence has concentrated on evaluating the first two goals: the proceeds from divestiture sales have greatly mitigated the amount of state-assisted stranded cost recovery, but divestiture--while restraining vertical market power--may have exacerbated horizontal market power. The induced reduction in generation capacity (making some IOUs net buyers of electricity generation), combined with a freeze on the retail prices charged to the end electricity users, effectively limited both the incentive and ability of incumbent IOUs to exercise vertical market power in wholesale electricity markets. (See Bushnell, Mansur, and Saravia (2005) and Mansur (2005).) But independent power producers who bought these divested plants have seemingly been able to exercise horizontal market power, exploiting both the lumpiness with which capacity was divested and constraints in the existing transmission network to drive up the relatively unconstrained wholesale prices. (See Borenstein, Bushnell, and Wolak (2002), Joskow and Kahn (2003), Puller (2007) and Wolak (1999).)

In contrast, there is little in the developing literature that evaluates the effectiveness of divestiture in achieving the more long-run goal of encouraging greater IPP entry and investment. Divestiture may be overall desirable if it helps foster beneficial long-run competition. Nominally, divestiture has increased IPP participation in U.S. wholesale electricity markets; many of the current major IPP participants participate through their ownership of divested IOU assets. However, whether divestiture led to greater IPP participation in real terms is unclear. Divestiture may have just "crowded out" new plant investments: acquisition of divested plants may have simply substituted for new plant investments by those same IPPs. The relevant comparison is not the actual comparison between IPP participation before and after divestiture but rather the counterfactual comparison between IPP participation in a market with divestiture and IPP participation in the same market but without divestiture. The main focus of this paper is the calculation of such a counterfactual.

In order to calculate this counterfactual comparison, a structural model of an IPP's power plant investment decision is proposed and estimated. Trying to study such a decision by means of cross-sectional regression on data from markets with and without divestiture is problematic as the presence of divestiture no doubt induces structural change in the IPP (reduced form) investment function. Divestiture not only changes the investment opportunity for an IPP but also the relationship between one IPP's investment decision and the investment decisions of its potential competitors: in order for an IPP to buy a divested power plant, the IPP must be willing to pay more for the divested asset than any of its competitors. As a consequence, there is a need to model explicitly how IPPs evaluate both the "making" of new power plants and the "buying" of existing, divested utility power plants.

The model adopted in this paper specifies the expected profit stream associated with a power plant, both new and old, and the investment cost associated with a new power plant as functions of exogenous plant, firm, market and regulatory variables. In a market without divestiture, an IPP invests when the expected profit stream from a new power plant is greater than the investment cost associated with the plant. (3) In a market with divestiture, an IPP must not only choose whether to invest but also how. This "make or buy" decision creates a link between an IPP's valuation of a new plant and the maximum amount an IPP is willing to pay for a divested plant: an IPP is willing to buy a plant as long as the value of the acquired plant (expected profit stream minus transaction price) is no less than either the value earned from building a new power plant or the value from making no investment. Assuming efficient divestiture sales, an IPP succeeds in buying a divested plant if the IPP has the highest willingness to pay among competitors. Similarly, an IPP builds a new plant if the IPP faces an expected profit stream from building a new plant that exceeds its investment cost and the opportunity cost of the capital. These revealed preference arguments provide the constraints on the data that identify the relevant expected profit and cost parameters.

The empirical model is applied to data on the investment decisions of 20 major IPPs from 1996 to 2000 for all 48 contiguous U.S. states. The estimated expected profit stream and investment cost functions provide some insights into observed IPP investment behavior. First, the estimates indicate a clear difference in the evaluation of power plants between IPPs affiliated and unaffiliated with electric utilities. Specifically, IPPs affiliated with electric utilities are found to face a greater investment cost associated with building a new plant and a modest profit advantage in running older plants. The result helps explain why the majority of divested plants have been bought by IOU-affiliated IPPs. Second, the estimates argue that the main incentive underlying the buying of power plants is the avoidance of the upfront investment cost associated with new plant construction, rather than the value associated with the (possibly desirable) location of old plants. Last, market characteristics reflecting the tightness of supply are found to have a significant, positive impact on expected profits in markets further along in restructuring. However, the estimates also find that, for markets with suitably adequate supply, further restructuring lowers expected profits.

The estimated expected profit stream and investment cost functions are used to calculate the counterfactual investment decisions of IPPs in the absence of divestiture. The counterfactuals show that, among the 32 (firm, market, year) observations in the sample where an IPP buys a divested power plant, on average only one would have resulted in the construction of a new power plant in the absence of divestiture. The simulations find that the total amount of new generation capacity "crowded out" by divestiture is very small, on average 177 megawatts (MW). This indicates that, while divestiture has not "crowded out" a large amount of new generation capacity, it has encouraged some new IPP participation in the restructured market.

The remainder of the paper is organized as follows. Section 2 provides an overview of generation investment since state-level electricity restructuring. Section 3 presents the theoretical model for the "make or buy" decision. Section 4 describes in detail the empirical analog to the theoretical framework. Section 5 presents and analyzes the parameter estimates of the model as well as the policy simulations based on the estimates. We conclude with some final thoughts.

2. Overview

* Since 1996, IPPs have had the opportunity to participate in some U.S. wholesale electricity markets without the explicit invitation of the incumbent electric utilities. (4) Figures 1 and 2 illustrate, geographically, the total generation investments made by the 20 IPPs in our sample from 1996 to 2001. (5) Figure 1 depicts the total for new power plant investments and Figure 2 the total for power plants acquired through divestiture. The sampled IPPs have invested in sizable (> 100 MW) generation capacity in 35 of the 48 contiguous states. Much of the investments are in the major restructured states. But some are in states that had not yet initiated restructuring, most notably in the Southeast. Furthermore, some divested utility plants were available to IPPs in "nonrestructured" states. These investments in nonrestructured states were spurred, in part, by the then conventional wisdom that restructuring would eventually occur nationwide: firms "jumping the gun" to establish an early market presence.

[FIGURE 1 OMITTED]

Table 1 shows that much of the electricity provided by IPPs early in the study period was generated from divested utility power plants. During our study period, the main divestiture concern was vertical market power: incumbent utilities could provide entrant IPPs unequal access to their downstream T&D network, thus raising their rivals' costs. Regulators felt that they could best circumvent this problem by having incumbent utilities largely exit the generation sector. The fact that these assets were sold in large lots, sometimes entire power systems to a single buyer, demonstrates the greater concern regulators placed on vertical than horizontal market power. (6) Consequently, divestiture allowed IPPs to acquire large capacity in restructured markets quickly.

Table 2 categorizes nondivestiture generation investments as capacity acquired in nonrestructured states, restructured states with no utility divestiture that year and restructured states with some utility divestiture that year. There are few new power plants built by IPPs in (state, year) where the acquisition of utility power plants is an option. This raises the concern that divestiture may be crowding out new power plant investments and underscores the need to investigate the impact of divestiture on IPP investment decisions.

[FIGURE 2 OMITTED]

We consider two main reasons why an IPP might prefer buying a divested power plant over developing a new one. First, divested plants may have characteristics unavailable to potential new plants. A prominent example is plant location. Given transmission constraints and the lack of available (comparable) industrial sites, a utility power plant may be situated in a location that makes the power plant particularly valuable in satisfying local electricity demand. (Fox (1999), Stavros (1999).) In this way, the potential revenue stream from an existing utility plant may be better than that of a new plant at a less desirable site. Second, buying divested plants may allow the IPP to avoid some of the investment costs associated with the development of new plants. The extent of this cost saving will depend on the price paid for the divested power plant. By buying an existing plant, the IPP may avoid some of the siting and permitting costs associated with developing a new plant.

However, buying a divested utility plant does not necessarily imply a "crowding out" of new plant investment. In the absence of divestiture, the IPP is not left with just the option of building a new plant; the IPP always has the "third" option of not investing at all. The lack of new investments in states with divestiture sales observed in Table 2 does not necessarily imply crowding out if such new investments were not economically viable. (7) Let {[V.sub.buy], [V.sub.make], [V.sub.none]} represent the value an IPP places on buying a divested plant, building a new plant and not investing in any plant, respectively. Observing an IPP buying a plant implies [V.sub.buy] [greater than or equal to] max{[V.sub.make], [V.sub.none]} but provides no information about the ordering of [V.sub.make] and [V.sub.none]. Crowding out happens only when [V.sub.make] [greater than or equal to] [V.sub.none]. Hence, observing an IPP buy a power plant is insufficient to tell us whether any new plant investment was crowded out. A study that investigates how an IPP evaluates both its new plant and divestiture investment options is necessary in order to investigate the counterfactual of what firms would have done in the absence of divestiture.

We propose and implement such a study. A key aspect of our chosen approach is its ability to distinguish between the two main benefits from buying a divested power plant: desirable plant characteristics and avoidance of some investment cost associated with new power plant development. We would expect "crowding out" to be of more concern if the main motive underlying the acquisition is the former--a firm can always avoid investment cost by not investing at all. Moreover, we do not base our study on the limited, published transaction prices for divestiture sales. Kahn (1999) discusses the cons of such an analysis. A theoretical model is developed that identifies how IPPs evaluate "make" and "buy" projects based on the most basic level of information: whether an IPP bought a divested plant, built a new plant or invested in no plants.

3. Theoretical model

* We assume that, in the same year and market, an IPP can carry out at most one type of positive investment: either acquisition of divested utility assets or the construction of new power plants. The idea is supported by our data: no firm is observed buying a divested asset and sinking a large amount of investment cost toward a new power plant in the same market and year. The idea is also consistent with the presence of some binding resource constraint that prevents the firm from exploring both "make" and "buy" options. (8) There is some evidence for such a constraint in the trade press, which discuss IPPs having difficulty raising capital through traditional means during the study period. (Stern (1998), Rigby (1999).) However, this constraint is neither modeled nor directly observed; it is simply imposed.

Given this basic assumption, we make the following specification concerning the firm's choice set and the timing of its decision. First, in a year when there is no divestiture in the market, the firm chooses between building a new plant and not investing at all. (9) Second, in a year when there is a divestiture sale in the market, the firm participates in the divestiture sale first. If the firm succeeds in buying the divested asset, it does not build a new plant in that market that year. The firm only explicitly considers building a new plant if it fails to buy any divested asset available in the market during that year. These assumptions lead us to the decision tree depicted in Figure 3.

When a firm participates in a divestiture sale, the firm factors in its option to build a new plant, an option which is forsaken if the firm is...



NOTE: All illustrations and photos have been removed from this article.



More articles from RAND Journal of Economics
Durable-goods oligopoly with secondary markets: the case of automobiles., 22-JUN-07
Nonlinear pricing in an oligopoly market: the case of specialty coffee, 22-JUN-07
Measuring consumer welfare in the CPU market: an application of the pure-characteristics demand model., 22-JUN-07

Looking for additional articles?
Click here to search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Click here to search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.