By Kennedy Maize
Construction Work in Progress (CWIP, pronounced “quip”) is a nasty spawn of utility accounting, encouraging sloppiness, sloth, and avarice in building practices. That’s the take in a new study by the conservative Manhattan Institute, “The Hidden Tax on Your Power Bill.”
“Construction Work in Progress, or CWIP, is the most influential energy policy term that never appears on a household utility bill,” the study says. It turns the way many utilities finance large construction projects, including utility investments, “on its head.”
Typically, the report notes, “A company raises debt and equity, builds the facility, and only after it is operational does it begin earning revenue. Investors shoulder the risk that costs may rise, schedules may slip, or demand may fall short. Customers start paying only once they receive the service.”
“Under CWIP, utilities face little pressure to control costs or timelines because every additional dollar spent and every month of delay expands their rate base and increases their guaranteed returns,” they said. “Cost overruns become profit opportunities rather than financial penalties.”
Many states with regulated investor-owned utilities, and federal government agencies including, prominently, the Federal Energy Regulatory Commission, routinely use it. FERC uses CWIP to provide an incentive for construction of interstate high-voltage electric transmission.
Former FERC commissioner and chairman Mark Christie long railed against CWIP. In a case last May, when the commission rejected a planned transmission project in Virginia and West Virginia and a bit of Maryland, Christie commented, “As I have said repeatedly over the past four years, it is long past time for this Commission to do its job of protecting consumers by cutting back on its unfair practice of handing out ‘FERC candy’ without any serious consideration of the impact on consumers already struggling to pay monthly power bills.”
The Manhattan Institute study explains that CWIP moves the costs of building a project into the owner’s “rate base,” which is used to determine what customers pay monthly on their bills. The companies earn a guaranteed rate of return including a profit on those bills. “As such” says the report, “when CWIP is approved, households are no longer just consumers of energy; they also become financiers, covering the carrying costs of projects still under construction.”
With CWIP, “the accumulating asset is included in the rate base. Accordingly, customer-service fees reflect the accumulated value of the incomplete (not in use) asset still under construction. I.e., the customer is paying for the new asset before it provides any service.” The results can be catastrophic.
The most egregious case of the damages of CWIP can be found in Georgia with the horror show that resulted when Georgia Power built two new 1000-MW nuclear power units in addition to two existing late 1980s-vintage nukes at its Vogtle station. When preliminary construction started on the two new units in 2009, the estimated cost was $14 billion. The two units went into service in 2023 and 2024 at a total cost of $37 billion.
Georgia’s legislature passed a law allowing CWIP in 2009, targeted directly for Vogtle. It allowed the utility to start including the costs of the job into consumer retail rates as soon as concrete was poured. The study remarked, “Customers effectively financed the project for 13 years before receiving a single kilowatt hour of new nuclear electricity.”
The study’s authors — Eric Olson and Jason M. Walter of the University of Tulsa and Jack Dorminey of West Virginia University — provided a concise and biting obituary of Vogtle 3 and 4: “Plant Vogtle’s trajectory makes the problems with CWIP vivid. Mounting construction setbacks from component redesigns, contractor bankruptcies, and pandemic-related labor shortages led to an explosion in the total price tag. Rather than face punishment, Georgia Power was entitled to earn over 10% return on that growing sum each year because CWIP was already in the rate base. For management and shareholders, the mushrooming budget did not threaten the project’s profitability; it enhanced it.”
Georgia’s voters this fall took some revenge for the Vogtle fiasco, although it didn’t return any unearned money to them. One of the members of the Georgia Public Service Commission who consistently bowed to Georgia Power, rejecting the advice of the professional staff, Republican Tim Echols, in November lost reelection to a Democratic consumer advocate. Echols had served on the commission since 2011.
CWIP appeared on the regulatory scene over a hundred years ago in Maryland in 1923, followed by Louisiana in 1924. According to the study, “No additional states permitted the practice from 1924 to 1947. After those two decades, the pace accelerated modestly, with eight states and the District of Columbia approving CWIP between 1947 and 1969, bringing the total to 11 jurisdictions by 1969.”
Then came nuclear power and with it an explosion of CWIP jurisdictions. According to the National Regulatory Research Institute, the use of CWIP nationwide grew from 8% of net electric utility plants in 1967 to 36% in 1983. Over a third of the net plant value of the industry was meaning that over one-third of the industry’s net plant value was tied up in ongoing construction.
According to the study, “That scale of capital encouraged overleveraging because utilities could finance larger portions of projects with debt while relying on CWIP funds to pay for the interest, reducing the market pressure to maintain conservative capital structures.”
Is reform possible as we appear to be heading into new boom times for large, expensive power plant construction projects, many tied to data center loads? For conventional projects, there are some accounting practices that could be viable, such as “allowance for funds used during construction” or AFUDC, CWIP’s kinder, gentler cousin. With AFUDC, notes the Manhattan Institute report, “Financing costs are separately capitalized as an asset. That asset is depreciated over its useful life, beginning after the new facility becomes operational.”
The advantage is that “existing customers do not bear the financing costs (via service fees) for an under construction project. But, importantly, it creates a different incentive structure for the utility.” The downside is rate shock when the plant goes into service. The choice of AFUDC is a basic step. Others could include making rates of return performance-based. Automatic clawback provisions when projects get into trouble could add discipline to the financing. Also, “for load pockets driven by data and AI campuses, an option is to move the load off the regulated grid entirely.”
“The complete elimination of CWIP in the rate base probably represents the best reform.” the scholars assert.