Katie Sieben

Dan Lipschultz   

Valerie Means  

Matthew Schuerger  

John Tuma  

In the Matter of a Commission Investigation to Identify and Develop Performance Metrics, and Potentially, Incentives for to Identify and Develop Performance Metrics, and Potentially, Incentives for Xcel Energy’s Electric Utility Operations






PUC Docket Number E-002/CI-17-40


The R Street Institute is a nonprofit, nonpartisan, free-market public policy organization based in Washington, D.C., that works to find market-based solutions to complex issues of public policy. We appreciated the opportunity to participate in the initial stakeholder meeting of this proceeding through our consultant, Chris Villarreal, and we look forward to continuing to engage as Phase I proceeds.

Keep It

The Commission and stakeholders should put the present undertaking into context. The point of devising metrics in the context of performance-based regulation (PBR) is to use them ultimately as part of a performance-incentive mechanism. The metrics selected in Phase I should eventually form a basis for Xcel’s compensation. R Street suggests those metrics should be few and focused—not more than two for each of the five outcomes the Commission has spelled out and as few as zero. We should be cognizant of the fact that the utility already possesses a strong incentive to accomplish certain outcomes, even while it lacks such incentives for others.[1]  

Minnesota is at the forefront of the movement to transform the utility cost-of-service paradigm. The Commission has set forth goals and outcomes for the utility to meet in accordance with recommendations from the Office of the Attorney General. These outcomes are affordability, reliability, customer service, environmental performance, and cost-effective alignment of generation and load. R Street recommends four metrics focused on two of those outcomes: affordability and alignment of generation and load. All four metrics are ‘new’ in that Xcel is not already required to report them to the Commission, and they are all easy to interpret and align with the Commission’s design principles.

Outcome: Affordability

PBR typically occurs within the context of a multiyear rate plan.
There are many advantages to establishing in advance the prices the utility is
allowed to charge and then allowing the utility to “go to work.” While
rate-making with shorter intervals between rate cases will introduce a bias for
capital, this type of rate-setting allows the company to more easily trade off
capital and operating expenditures. It also creates incentives for a utility to
be creative, more customer-friendly, and open to third-party solutions to its
operational challenges.

The existing regime of pass-through and tracker mechanisms can be at odds with the core purposes of multiyear rate plans. Xcel currently employs approximately a dozen trackers to recover costs between rate cases, encompassing the costs of fuel, transmission, conservation, and other environmental spending.[2] By allowing a utility to recover a significant portion of its total costs through accounting deferrals, regulation can diminish the incentive to create efficiencies around these operating expenditures. There may be good reasons to have trackers, but because some of these costs may be largely outside the utility’s control, there should be a performance consideration directed toward affordability that takes a global perspective of how the utility costs that customers are required to pay are changing over the multiyear period.

We therefore offer two potential metrics that could be used to
capture this phenomenon and recommend that at least one of them be adopted for
the Commission’s affordability outcome.

  1. Utility Prices versus Economy-Wide Measures: This metric would compare the utility’s prices, under a multiyear rate plan with trackers, to measurements of customer costs in the wider economy. Since a utility exists to serve the economy as a whole and is not a consumer end in itself, a metric that ties its price performance to other parts of the consumer economy is appropriate. We propose establishing a ratio where the annual percentage change in Xcel rates for residential and commercial customers is the numerator and an economy-wide measure is the denominator. The denominator should be a percentage that is the difference between inflation and productivity, where inflation is the Consumer Price Index and productivity is non-farm labor productivity measured by the U.S. Department of Labor. A ratio that is in excess of 1.0 would suggest that the utility’s prices are increasing more quickly than prices within the wider economy, while a metric of less than 1.0 would suggest the opposite.[3] This metric, known as “GPI – X,” is sometimes the basis of an adjustment to rates year-on-year within a multiyear plan. However, it is appropriate to use this metric to measure affordability outcomes over the course of a multiyear rate plan because of the potentially substantial influence of trackers on the costs consumers must pay. This metric, if performance around it was compensable, would provide a stronger incentive to the utility to control overall costs—including by seeking out efficiencies that may exist in fuel costs and wholesale electricity markets. It also makes intuitive sense that a firm that holds its prices below economy-wide measures should enjoy a higher return, and vice versa. The data is easily verifiable and relies on published third-party sources. It aligns with the overarching policy goal of service at just and reasonable rates, and it also supports environmentally positive outcomes by providing a utility with an incentive to control fuel costs.
  2. Tracker Revenue as a Proportion of Total Revenue: Another way to ascertain the impact of pass-through and tracker mechanisms on customer affordability is to measure their effect more directly. Here, the metric we propose would be a ratio where the numerator is the total operating revenue garnered from trackers and the denominator is the utility’s total revenue. A ratio that increases year on year suggests more costs are shifting to customers through trackers and, concomitantly, that the utility’s business risk associated with under-recovery of costs is diminishing. Like the first metric we propose, it is appropriate that a utility’s returns on equity be adjusted to account for how much revenue it obtains through pass-through mechanisms, as opposed to collections that are made through rates that do not change automatically on the basis of actual costs. This metric, like the “GPI – X” metric discussed above, serves to mitigate the information asymmetry concerning internal operating costs that exists between the regulated company and the regulator because it takes a perspective on total costs. It also mitigates the perverse incentive that exists when utilities receive little or no benefits from better management of operating costs—something other aspects of PBR are specifically designed to overcome.[4] The metric is clearly defined and linked to the policy goal of greater operational efficiency. The data is easily verifiable, quantifiable, and does not interfere with existing regulations.

Outcome: Cost-Effective
Alignment of Generation and Load

While the underlying purpose of this outcome was
discussed with some skepticism at the last stakeholder meeting, we believe it
is actually one of the most important outcomes. Xcel is embarking on a
transformation of its power-supply portfolio. Many of the resources in which it
invests will be intermittent, and the value of its invested capital will
ultimately be in how those generators align—or do not align—to both customer
demand and to the wider wholesale market for electricity. It is therefore
important that the Commission adopt metrics that measure the value of
generation. One can understand such metrics as being a PBR associated with the
integrated resource-planning process. Currently, the Commission embarks on an ex
anteevaluation of Xcel’s resource planning.
Metrics on this outcome would allow the Commission to see, and reward, how
resource planning and acquisition actually performs. It is important to align Xcel’s
investments cost-effectively to load, maximizing the value of renewables and
ultimately lowering the cost to customers of a clean-energy transition.

As with affordability—and even though a single, well-designed
metric may be sufficient to ascertain the cost-effective alignment of
generation and load—we propose two metrics.

  1. Portfolio Value to Customers: One way to understand whether the Xcel portfolio is well-matched to its customers’ demands is to create a metric that measures the portfolio’s performance relative to the market. For this purpose, we propose two ratios. The denominator would be the average energy price at the Midcontinent Independent System Operator’s (MISO) Minnesota Hub over the previous year.[5] The numerators would be the average energy price Xcel is paid (or pays) per megawatt-hour sold into (or purchased from) the wholesale market. A ratio that is, year on year, increasing for sales (or decreasing for purchases) suggests that Xcel’s generation is becoming more cost-effectively aligned with load. Put another way, this metric measures whether the resources that Xcel has rate-based are actually valuable relative to objective measurements made about value that emerge through MISO’s energy prices.[6] This metric also tacitly encompasses proposed metrics—suggested by stakeholders—concerning load shape and load shifting, which affect the net positions that drive Xcel’s interactions with the wider wholesale market. This metric could be tailored toward peak or shoulder hours, or to certain months of the year, or to particular nodal prices, rather than to the Minnesota Hub. However, we believe it is best to provide a broad, system-wide measurement to ensure that it can be measured on a portfolio basis. This metric benefits from using data that is routinely produced and that can be verified by third parties.
  2. Peak Reduction: This metric would measure system peak, with the intention of rewarding reductions after a multiyear interval.[7] Proactive efforts to reduce peak demand can generate significant reductions in customer costs by shifting peak energy demand to lower-cost hours, reducing transmission charges incurred during peak demand, and reducing installed capacity requirements over time. A systemic measurement like this one is superior to metrics that measure the growth of programs that are means to an end of peak reductions because it properly measures outcomes and not inputs. A normalizing adjustment for weather would be appropriate for this metric, and while this adds some complexity, it is a complexity that Minnesota already must attend to in its ratemaking conventions. Like the other metrics we propose, this one adheres to the Commission’s desire for easy interpretation, reliably available data, and metrics and data that can be verified by third parties.


A move away from a capital bias is implicit in the movement away from a strict cost-of-service regulatory model and toward the PBR model.Xcel makes the point in a previous filing that a multiyear rate plan itself is a form of PBR because the locking-in of rates is a tool used to strengthen a utility’s incentives to contain costs.[8] We fully agree. However, we believe that the metrics discussed above are useful and necessary complements to a multiyear rate plan which, because of both trackers and long-lived capital assets, will not alone provide a full complement of incentives to align with either the public interest or Minnesota’s public policy.

Overly specific direction and the promotion of highly specific outcomes
should be avoided. As an example, greater electrification in transportation and
home heating can be beneficial to the environment and personal well-being, but these
outcomes should not be prescribed as metrics. A well-regulated utility under a
multiyear rate plan will already have a strong incentive to engage in
beneficial electrification.

R Street welcomes this opportunity to participate in Minnesota’s
efforts on this important topic and thanks the Commission for its consideration
of our comments.

Respectfully submitted,

William Murray
Energy Policy Manager

The R Street Institute
1212 New York Ave. N.W.,
Suite 900
Washington, D.C. 20005
[email protected]

Travis Kavulla
Energy Policy Director

The R Street Institute
1212 New York Ave. N.W.,
Suite 900
Washington, D.C. 20005
[email protected]

One of the most sophisticated PBR proposals made recently in the United States
provides additional compensation based on only four metrics. Testimony of H.O.
Pieper, Massachusetts Department of Public Utilities, “Petition of
Massachusetts Electric Company and Nantucket Electric Company,” Docket No.
18-150, Nov. 15, 2018, p. 54.

Xcel Energy, “Rate Riders.”

This assumes both the numerator and denominator are positive values. A
provision would need to be made when one or both of these values is negative.

Kenneth W. Costello, “A Model Multiyear Rate Plan for Public Utilities,” University
of Illinois: Frameworks for Regulation of Public Utilities in the 21st Century,
February 2019, p. 28.

The real-time Minnesota Hub pricing, together with more granular locational marginal
prices, can be found online. See MISO Energy, “Contour Map.”

While of course MISO’s prices at the Minnesota Hub are not within the exclusive
control of Xcel’s activities, the use of that data for this metric is really
intended to be a yardstick against which the value of Xcel’s resources is measured.

Stakeholders may look to the National Gridproceeding in Massachusetts for an example of a proposed metric and
performance-incentive mechanism associated with peak reductions. See generally
Testimony of H.O. Pieper.

[8] “Comments
of Xcel Energy In the Matter of the
Commission Investigation to Identify and Develop Performance Metrics and
Potentially, Incentives for Xcel Energy’s Electric Utility Operations,”
Minnesota Public Utilities Commission, Docket No. E002/CI-17-401, Dec. 21, 2017,
pp. 6, 14.

Featured Publications