• Selected
• Entire List

Abstract:
Understanding the sensitivity of gasoline demand to changes in prices and income has important implications for policies related to climate change, optimal taxation and national security, to name only a few. While the short-run price and income elasticities of gasoline demand in the United States have been studied extensively, the vast majority of these studies focus on consumer behavior in the 1970s and 1980s. There are a number of reasons to believe that current demand elasticities differ from these previous periods, as transportation analysts have hypothesized that behavioral and structural factors over the past several decades have changed the responsiveness of U.S. consumers to changes in gasoline prices. In this paper, we compare the price and income elasticities of gasoline demand in two periods of similarly high prices from 1975 to 1980 and 2001 to 2006. The short-run price elasticities differ considerably: and range from -0.034 to -0.077 during 2001 to 2006, versus -0.21 to -0.34 for 1975 to 1980. The estimated short-run income elasticities range from 0.21 to 0.75 and when estimated with the same models are not significantly different between the two periods.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
Understanding the sensitivity of gasoline demand to changes in prices and income has important implications for policies related to climate change, optimal taxation and national security, to name only a few. While the short-run price and income elasticities of gasoline demand in the United States have been studied extensively, the vast majority of these studies focus on consumer behavior in the 1970s and 1980s. There are a number of reasons to believe that current demand elasticities differ from these previous periods, as transportation analysts have hypothesized that behavioral and structural factors over the past several decades have changed the responsiveness of U.S. consumers to changes in gasoline prices. In this paper, we compare the price and income elasticities of gasoline demand in two periods of similarly high prices from 1975 to 1980 and 2001 to 2006. The short-run price elasticities differ considerably and range from -0.034 to -0.077 during 2001 to 2006, versus -0.21 to -0.34 for 1975 to 1980. The estimated short-run income elasticities range from 0.21 to 0.75 and when estimated with the same models are not significantly different between the two periods. One implication of these findings is that gasoline taxes would need to be significantly larger today in order to achieve an equivalent reduction in gasoline consumption. This, coupled with the political difficulties in adopting gasoline taxes, suggests that policies and technologies designed to improve fuel economy are likely becoming relatively more attractive as a means to reduce fuel consumption.

Gasoline demand, Price elasticity

Abstract:
In this paper, we present an empirical analysis of deregulated electricity prices. We begin by examining the distributional and temporal properties of the price process in a non-parametric framework. This analysis is followed by comparing the forecasting ability of several different statistical models. The findings reveal several characteristics unique to electricity prices, including deterministic components of the series at different frequencies and a high degree of persistence in the price level. An "inverse leverage effect" is also found, where positive shocks to the price series result in larger increases in volatility than negative shocks. Results consistent with other asset prices, such as time-varying volatility are also uncovered. We find that existing financial models of asset prices fail to forecast the extremely erratic nature of electricity prices. Non-Markovian specifications, in conjunction with exogenous information (e.g. weather), are a necessary starting point for practical applications, such as security pricing.

Abstract:
Electricity and natural gas markets have traditionally been serviced by one of two market structures. In some markets, electricity and natural gas are sold by a dual-product regulated monopolist, while in other markets, electricity and natural gas are sold by separate single-product regulated monopolies. This paper analyzes the relative pricing and investment decisions of electricity firms operating in the two market structures. The unique relationship between these two products, namely that electricity and natural gas are substitutes in consumption and natural gas is an input into the generation of electricity, allows me to gain inferences regarding the efficacy of regulation in both the electricity and natural gas industries. The results imply that both electricity prices and reliance on natural gas generation are higher in a dual-product setting, both suggestive that regulators respond to the relative incentives of electricity and natural gas firms.

Abstract:
Many policy-makers are currently weighing the advantages of deregulating electricity markets over more traditional regulatory methods. However, within this traditional regulatory environment many options exist. In particular, the use of incentive regulation programs in US electricity markets has grown during the past two decades. These programs differ in both their goals and how they attempt to meet these goals. In this paper, I discuss the wide array of programs that have been utilized, and investigate the impact of individual programs on the technical efficiency of a large set of coal and natural gas generator units. Within a stochastic frontier framework, I allow the distribution of inefficient production to be a function of the regulatory environment the plant operates under. The results suggest that while certain incentive regulations increase observed technical efficiency, others have either no effect or even lead to a reduction in efficiency.

Abstract:
While the electricity industry is currently experiencing a regime shift from state regulation to competitive markets for generation, a number of questions regarding the change in regulatory environment in the electricity industry during the beginning of this century remain unsettled. This paper revisits some of these issues. Specifically, the paper tests the validity of the two most commonly suggested reasons for the origin of state regulation in the electricity industry. Using Census data gathered on the electricity industry during the beginning of this century, the paper models the decision to adopt state regulation as a function of the average electricity price and profit rate in the state. The results cast doubt on the commonly accepted justification that state regulation was passed to limit competition and increase the profits of electricity firms, but do not unilaterally support the public interest view, either. The paper then draws on trade publications to assess the industry's mindset during this period. Literature from the National Electric Light Association suggests that although some industry leaders supported state regulation in order to rid themselves of corrupt local politicians, the industry did not unilaterally support increases in regulation.

Abstract:
We test whether a non-binding price ceiling may serve as a focal point for tacit collusion, using data from the credit card market during the 1980s. In our sample, most credit card issuers face a state-level interest rate ceiling, and well over half match their ceiling. We develop an empirical model that can separately identify the instance in which an issuer matches its ceiling because it is binding, and the instance in which an issuer matches its ceiling even though it is not binding. The model yields evidence in favor of tacit collusion: a statistically significant proportion of issuers match their ceiling even though it is not binding. Within a state, tacit collusion is less likely as the ceiling rises, more likely as concentration or costs rise, and less likely in periods of high demand. We also find that entry into credit cards is higher where we find evidence of tacit collusion, and lower where we find evidence that a ceiling is binding. It appears that tacit collusion became less prevalent over the 1980s, as entry into credit cards surged nationwide. The results highlight a perverse effect of price cap regulation.

Focal Points, Tacit Collusion, Price Ceilings, Double Hurdle

Abstract:
Incompatibility in markets with indirect network effects can affect prices if consumers value "mix and match" combinations of complementary network components. In this paper, we examine the effects of incompatibility using data from a classic market with indirect network effects: Automated Teller Machines (ATMs). Our sample covers a period during which higher ATM fees increased incompatibility between ATM cards (which are bundled with deposit accounts) and other banks' ATM machines. A series of hedonic regressions suggests that incompatibility strengthens the relationship between deposit account pricing and own ATMs, and weakens the relationship between deposit account pricing and competitors' ATMs. The effects of incompatibility are stronger in areas with high population density, suggesting that high travel costs increase both the strength of network effects and the importance of incompatibility in ATM markets.

Abstract:
Incompatibility in markets with indirect network effects can reduce consumers' willingness to pay if they value "mix and match" combinations of complementary network components. For integrated firms selling complementary components, incompatibility should also strengthen the demand-side link between components. In this paper, we examine the effects of incompatibility using data from a classic market with indirect network effects: Automated Teller Machines (ATMs). Our sample covers a period during which higher ATM fees increased incompatibility between ATM cards and other banks' ATM machines. We find that incompatibility led to lower willingness to pay for deposit accounts. We also find that incompatibility benefited firms with large ATM fleets.

Abstract:
A low carbon fuel standard (LCFS) seeks to reduce greenhouse gas emissions by limiting a fuel producer's carbon emissions per unit of output. California has launched an LCFS for transportation fuels; others have called for a national LCFS. We show that this policy decreases production of high-carbon fuels but increases production of low-carbon fuels. The net effect of this may be an increase in carbon emissions. The LCFS cannot be first best, and the best LCFS may reduce social welfare. We simulate the outcomes of a national LCFS, focusing on gasoline and ethanol as the high- and low-carbon fuels. For a broad range of parameters, we find that the LCFS is unlikely to increase CO2 emissions. However, the surplus losses from the LCFS are likely to be quite large ($80 to $760 billion annually for a national LCFS reducing carbon intensities by 10 percent), energy prices are likely to increase, and the average carbon cost ($307 to $2,272 per ton of CO2 for the same LCFS) can be much larger than damage estimates. We describe an efficient policy that achieves the same emissions reduction at a much lower surplus cost ($16 to $290 billion) and much lower average carbon cost ($60 to $868 per ton of CO2).

Abstract:
A low carbon fuel standard (LCFS) seeks to reduce greenhouse gas emissions by capping an industry's carbon emissions per unit of output. California has launched an LCFS for automotive fuels; others have called for a national LCFS. We show that this policy causes production of high-carbon fuels to decrease but production of low-carbon fuels to increase. The net effect of this may be an increase in carbon emissions. The LCFS may also reduce welfare, and the best LCFS may be no LCFS. We simulate the outcomes of a national LCFS, focusing on gasoline and ethanol as the high- and low-carbon fuels. For a broad range of parameters, we find that the LCFS is unlikely to increase CO2 emissions. However, the surplus losses from the LCFS are quite large ($80 to $760 billion annually for a national LCFS reducing carbon intensities by 10 percent), and the average carbon cost ($307 to $2,272 per ton of CO2 for the same LCFS) can be much larger than damage estimates. We propose an efficient policy that achieves the same emissions reduction at a much lower surplus cost ($16 to $290 billion) and much lower average carbon cost ($60 to $868 per ton of CO2).

carbon, externalities, pollution, taxes, transportation, gasoline, ethanol

Abstract:
Estimating market power is often complicated by the lack of reliable measures of marginal cost. Instead, policy-makers often rely on other summary statistics of the market, thought to be correlated with price cost margins - such as concentration ratios or the HHI. In many industries, these summary statistics may be only weakly correlated with deviations from perfectly competitive pricing. Beginning with Gollop and Roberts (1979), a number of empirical studies have allowed the data to identify industry competition and marginal cost levels by estimating the firms' first order condition within a conjectural variations framework. Despite the prevalence of such New Empirical Industrial Organization (NEIO) studies, Corts (1999) illustrates the estimated mark-up levels may be biased, since the estimated conjectural variations model forces the supply relationship to be a ray through the marginal cost intercept, whereas this need not be true in dynamic games. In this paper, we use direct measures of marginal cost for the California electricity market to measure the extent to which estimated mark-ups and marginal costs are biased. Our results suggest that the NEIO technique poorly estimates the level of mark-ups and the sensitivity of marginal cost to cost shifters.

NEIO, Market Power, Electricity

Abstract:
While the Telecommunications Act of 1996 introduces competition into local telephone markets at a national level, prior to the Act many states had already restructured their regulatory methods. In particular, many states allowed for competition in both local service and "local long distance" markets prior to 1996. In addition to the expansion of competition, there has been an increased use of price regulation (price caps and price freezes), rather than traditional cost based regulatory methods in determining and updating rates for local exchange and local long distance toll services. Using a three-dimensional panel data set, I investigate whether price regulation and differences in entry conditions in local exchange and local long distance toll markets have had an impact on rate changes for residential local service, business local service, and local long distance toll service for the incumbent local exchange companies. In doing so, I estimate an empirical model that accounts for the stickiness of rates in regulated industries. The model, based on (S,s) models that have developed in a number of literatures, treats firms as maximizing expected profits in the price adjustment costs and regulatory uncertainty resulting from rate hearings, and regulators as controlling profits in the presence of adjustment costs. I find that competition has prompted a significant amount of rate rebalancing that has reduced the amount of cross-subsidization present in these markets. In addition, the findings suggest that price cap regulation leads to higher rates, relative to rate-of-return regulation.

Abstract:
Incompatibility in market with network effects reduces consumers' ability to "mix and match" components offered by different sellers, but can also spur changes in product attributes that might benefit consumers. In this paper, we estimate the effects of incompatibility on consumers in a classic hardware/software market: ATM cards and machines. We find that while ATM fees ceteris paribus reduce the network benefit from other banks' ATMs, a surge in ATM deployment accompanies the shift to surcharging. This is valuable to consumers and often completely offsets the harm from higher fees. The results suggest that policy discussions of incompatibility must consider not only its direct effect on consumers, but also its effect on product attributes.

Abstract:
Incompatibility in markets with network effects can either benefit or harm consumers. Incompatibility reduces consumers' ability to "mix and match" components offered by different sellers, but can also be associated with changes in product attributes that might benefit consumers. In this paper, we estimate the effects of incompatibility in a classic hardware/software market: ATM cards and machines. Our empirical model allows us to measure the indirect network effect relating the value of ATM cards to ATM availability. It also allows us to measure the effects of incompatibility as measured by ATM fees. Our sample contains a relatively discrete move toward incompatibility after 1996, when banks began to impose surcharges on non-customers using their ATM machines. We provide estimates of the partial equilibrium effects of increased incompatibility on consumer welfare, finding that ATM fees ceteris paribus reduce the indirect network effect associated with other banks' ATMs. However, a surge in ATM deployment accompanies the shift to surcharging and in many cases completely offsets the reduction in welfare associated with higher fees. This suggests that welfare analyses should consider the interaction between incompatibility and changes in product attributes.

Abstract:
We test whether firms use incompatibility strategically, using data from ATM markets. High ATM fees degrade the value of competitors' deposit accounts, and can in principle serve as a mechanism for siphoning depositors away from competitors or for creating deposit account differentiation. Our empirical framework can empirically distinguish surcharging motivated by this strategic concern from surcharging that simply maximizes ATM profit considered as a standalone operation. The results are consistent with such behavior by large banks, but not by small banks. For large banks, the effect of incompatibility seems to operate through higher deposit account fees rather than increased deposit account base.

Abstract:
We test whether firms use incompatibility strategically, using data from ATM markets. High ATM fees degrade the value of competitors' deposit accounts, and can in principle serve as a mechanism for siphoning depositors away from competitors or for creating deposit account differentiation. Our empirical framework can empirically distinguish surcharging motivated by this strategic concern from surcharging that simply maximizes ATM profit considered as a stand-alone operation. The results are consistent with such behavior by large banks, but not by small banks. For large banks, the effect of incompatibility seems to operate through higher deposit account fees rather than increased deposit account base.

Abstract:
Estimating market power is often complicated by a lack of reliable marginal cost data. A number of empirical studies identify industry competition and marginal cost levels by estimating the firms' first order condition within a conjectural variations framework. Few studies, however, have analyzed the accuracy of this technique. In this paper, we use direct measures of marginal cost for the California electricity market to measure the extent to which estimated mark-ups and marginal costs are biased. Our results suggest that the technique poorly estimates mark-ups and the sensitivity of marginal cost to cost shifters.

Abstract:
Estimating market power is often complicated by the lack of reliable measures of marginal cost. Instead, policy-makers often rely on other summary statistics of the market, thought to be correlated with price cost margins---such as concentration ratios or the HHI. In many industries, these summary statistics may be only weakly correlated with deviations from perfectly competitive pricing. Beginning with Gollop and Roberts (1979), a number of empirical studies have allowed the data to identify industry competition and marginal cost levels by estimating the firms' first order condition within a conjectural variations framework. Despite the prevalence of such "New Empirical Industrial Organization" (NEIO) studies, Corts (1999) illustrates the estimated mark-up levels may be biased, since the estimated conjectural variations model forces the supply relationship to be a ray through the marginal cost intercept, whereas this need not be true in dynamic games. In this paper, we use direct measures of marginal cost for the California electricity market to measure the extent to which estimated mark-ups and marginal costs are biased. Our results suggest that the NEIO technique poorly estimates the level of mark-ups and the sensitivity of marginal cost to cost shifters.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
We study price convergence between the two major markets for wholesale electricity in California from their deregulation in April 1998 through November 2000, nearly the end of trading in one market. We would expect profit-maximizing traders to have eliminated persistent price differences between the markets. Institutional impediments and traders' incomplete understanding of the markets, however, could have delayed or prevented price convergence. We find that the two benchmark electricity prices in California -- the Power Exchange's day-ahead price and the Independent System Operator's real-time price -- differed substantially after the markets opened but then appeared to be converging by the beginning of 2000. Starting in May 2000, however, price levels and price differences increased dramatically. We consider several explanations for the significant price differences and conclude that rapidly changing market rules and market fundamentals, including one buyer's attempt to exercise a form of monopsony power, made it difficult for traders to take advantage of opportunities that ex post appear to have been profitable.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
US Electricity and natural gas markets have traditionally been serviced by one of two market structures. In some markets, electricity and natural gas are sold by a regulated dual-product monopolist, while in other markets, electricity and natural gas are sold by separate regulated single-product monopolies. I analyze whether electricity and natural gas prices depend on the market structure and compare these results to the predictions of a number of theories. The results are most consistent with the political economy theories suggesting that regulators respond to interest group activity.

Abstract:
In the second-half of the 1990s, the positive impact of information technology on productivity growth for the United States became apparent. The measurement of this productivity improvement depends on hedonic procedures adopted by the Bureau of Labor Statistics (BLS) and Bureau of Economic Analysis (BEA). In this paper we suggest a new reason why conventional hedonic methods may overstate the price decline of personal computers. We model computers as a durable good and suppose that software changes over time, which influences the efficiency of a computer. Anticipating future increases in software, purchasers may "overbuy" characteristics, in the sense that the purchased bundle of characteristics is not fully utilized in the first months or year that a computer is owned. In this case, we argue that hedonic procedures do not provide valid bounds on the true price of computer services at the time the machine is purchased with the concurrent level of software. To assess these theoretical results we estimate the model and find that before 2000 the hedonic price index constructed with BLS methods overstates the fall in computer prices. After 2000, however, the BLS hedonic index falls more slowly, reflecting the reduced marginal cost of acquiring (and therefore marginal benefit to users) of characteristics such as RAM, hard disk space or speed.

Abstract:
We analyze tacit collusion in an industry characterized by cyclical demand and long-run scale decisions; firms face deterministic demand cycles and choose capacity levels prior to competing in prices. Our focus is on the nature of prices. We find that two types of price wars may exist. In one, collusion can involve periods of mixed strategy price wars. In the other, consistent with the Rotemberg and Saloner (1986) definition of price wars, we show that collusive prices can also become countercyclical. We also establish pricing patterns with respect to the relative prices in booms and recessions. If the marginal cost of capacity is high enough, holding current demand constant, prices in the boom will be generally lower than the prices in the recession; this reverses the results of Haltiwanger and Harrington (1991). In contrast, if the marginal cost of capacity is low enough, then prices in the boom will be generally higher than the prices in the recession. For costs in an intermediate range, numerical examples are calculated to show specific pricing patterns.

Abstract:
This paper examines the adoption of state electricity regulation around the beginning of the 20th century. I model this decision as a hazard rate to determine what influenced the adoption of state regulation. I find that adoption is positively correlated with capacity shortages, greater wealth and lower residential electricity penetration rates. These results suggest that state regulation responded to regulatory inefficiencies and residential consumer interests. In addition, adoption rates were higher in states that had a strong industrial and coal mining presence. These results are consistent with the interest group and contracting theories of regulation.

Abstract:
For political and practical reasons, environmental regulations sometimes treat point source polluters, such as power plants, differently from mobile source polluters, such as vehicles. This paper measures the extent of this regulatory asymmetry in the case of nitrogen oxides (NOx), the criteria air pollutant that has proven to be the most recalcitrant in the United States. We find significant differences in marginal abatement costs across source types with the marginal cost of reducing NOx from cars less than half of the marginal cost of reducing NOx from power plants. Our findings have important implications for the efficiency of NOx emissions reductions and, more broadly, the benefits from increasing the sectoral scope of environmental regulation. We estimate that the costs of achieving the desired emissions reductions could have been reduced by nearly $2 billion, or 9 percent of program costs, had marginal abatement costs been equated across source types.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
Empirical exercises in economics frequently involve estimation of highly nonlinear models. The criterion function may not be globally concave or convex and exhibit many local extrema. Choosing among these local extrema is non-trivial for a variety of reasons. In this paper, we analyze the sensitivity of parameter estimates, and most importantly of economic variables of interest, to both starting values and the type of non-linear optimization algorithm employed. We focus on a class of demand models for differentiated products that have been used extensively in industrial organization, and more recently in public and labor. We find that convergence may occur at a number of local extrema, at saddles and in regions of the objective function where the first-order conditions are not satisfied. We find own- and cross-price elasticities that differ by a factor of over 100 depending on the set of candidate parameter estimates. In an attempt to evaluate the welfare effects of a change in an industry's structure, we undertake a hypothetical merger exercise. Our calculations indicate consumer welfare effects can vary between positive values to negative seventy billion dollars depending on the set of parameter estimates used.

Abstract:
The dramatic increase in gasoline prices from close to $1 in 1999 to $4 at their peak in 2008 made it much more expensive for consumers to operate an automobile. In this paper we investigate whether consumers have adjusted to gasoline price changes by altering what automobiles they purchase and what prices they pay. We investigate these effects in both new and used car markets. We find that a $1 increase in gasoline price changes the market shares of the most and least fuel-efficient quartiles of new cars by +20% and -24%, respectively. In contrast, the same gasoline price increase changes the market shares of the most and least fuel-efficient quartiles of used cars by only +3% and -7%, respectively. We find that changes in gasoline prices also change the relative prices of cars in the most fuel-efficient quartile and cars in the least fuel-efficient quartile: for new cars the relative price increase for fuel-efficient cars is $363 for a $1 increase in gas prices; for used cars it is $2839. Hence the adjustment of equilibrium market shares and prices in response to changes in usage cost varies dramatically between new and used markets. In the new car market, the adjustment is primarily in market shares, while in the used car market, the adjustment is primarily in prices. We argue that the difference in how gasoline costs affect new and used automobile markets can be explained by differences in the supply characteristics of new and used cars.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
New car fleet fuel economy, weight and engine power have changed drastically since 1980. These changes represent both movements along and shifts in the fuel economy/weight/engine power production possibilities frontier. This paper estimates the technological progress that has occurred since 1980 and the trade-offs that manufacturers and consumers face when choosing between fuel economy, weight and engine power characteristics. The results suggest that if weight, horsepower and torque were held at their 1980 levels, fuel economy for both passenger cars and light trucks could have increased by nearly 50 percent from 1980 to 2006; this is in stark contrast to the 15 percent by which fuel economy actually increased. I also find that once technological progress is considered, meeting the CAFE standards adopted in 2007 will require halting the observed increases in weight and engine power characteristics, but little more; in contrast, the standards recently announced by the new administration, while certainly attainable, require non-trivial downsizing. I also investigate the relative efficiencies of manufacturers. I find that US manufacturers tend to be above the median in terms of their passenger vehicle fuel efficiency conditional on weight and engine power, and are among the top for light duty trucks; Honda is the most efficient manufacturer for both passenger cars, while Volvo is the most efficient manufacturer of light duty trucks. However, I also find that over time, US manufacturers' relative efficiency in both passenger cars and light trucks has degraded. These results may provide insight into their current financial troubles.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
To determine whether the investment decisions of institutional plan sponsors contribute to their asset values, this study used a dataset of 80,000 yearly observations of institutional investment product assets, accounts, and returns for 1984−2007. Results show that plan sponsors may not be acting in their stakeholders’ best interests when they make rebalancing or reallocation decisions. Investment products that receive contributions subsequently underperform products experiencing withdrawals over one, three, and five years. For investment decisions among equity, fixed-income, and balanced products, most of the underperformance can be attributed to product selection. Tests suggest that these results are not attributable to survivorship or other biases. Much like individual investors who switch mutual funds at the wrong time, institutional investors do not appear to create value from their investment decisions.

Performance Measurement and Evaluation, Performance Attribution, Manager Selection, Portfolio Management, Asset Allocation, Portfolio Construction, Rebalancing and Implementation

Abstract:
If consumers value ‘mix and match’ combinations of network complements, incompatibility between different sellers' components should affect prices. In ATM markets, a 1996 governance change exogenously generated such incompatibility, by allowing banks to impose surcharges when other banks' deposit customers use their ATM's. In our data, incompatibility makes the relationship between deposit account pricing and own ATM's more positive, and makes the relationship between deposit account pricing and competitors ATM's more negative. The level effect on prices is positive. The pattern of results is more pronounced in high population density markets, where customers may care more about ATM's.

Abstract:
For two years prior to the collapse of California's restructured electricity market, power traded in both a forward and a spot market for delivery at the same times and locations. Nonetheless, prices in the two markets often differed in significant and predictable ways. This apparent inefficiency persisted, we argue, because most firms believed that trading on inter-market price differences would yield regulatory penalties. For the few firms that did make such trades, it was not profit-maximizing to eliminate the price differences entirely. Skyrocketing prices in 2000 changed the major buyers' (utilities') incentives and exacerbated the price differentials between the markets.