Exit, Voice, and Civil Disobedience

The political decentralization of the grid

“There is one great advantage that must follow regulation, and that advantage is protection”

– Samuell Insull

“You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.”

– Buckminister Fuller

A few months ago, the renowned venture capital fund a16z broke into the distributed energy scene and revealed it is DER-pilled. Their first formal foray highlighted all the familiar techno-economic maladies accelerating the decentralization of the electricity grid, including bloated interconnection processes, rapidly rising delivery costs, and growing reliability concerns. It then took stock of a smattering of enabling technologies before dropping the hammer and concluding that the centralized grid “will be consumed by networks of decentralized assets growing around it.” 

VCs make a living by evaluating big-picture economic and technological trends. But for many entrepreneurs slogging through an endless morass of utility pilot programs or seeing their entire business models threatened by the opaque and seemingly arbitrary decisions of state Public Utility Commissions (PUCs), it is a struggle to see how this grand vision – one that I and many others generally support – can come to fruition when confronted with the messy reality of how the grid is governed. For how can we unlock the fast iteration and experimentation necessary to disrupt a system embroidered with a thick layer of state regulatory institutions? 

The purpose of this essay, and those to come, is to provide increasingly higher fidelity visions of how political decentralization of the grid, or rather re-decentralization, can evolve in concert with the technical decentralization of the grid that a16z highlighted. It involves looking closely at the mechanisms by which institutions evolve in concert with culture and technology, and the best high-profile case study to build this lens takes us to what happened in Maine last fall. 

In October 2023, in response to decades of rising costs, outages, and poor service, Mainers took to the ballot with the hopes of reclaiming their electrical grid infrastructure. The issue at hand was whether to revoke the franchise licenses for Central Maine Power and Versant – two foreign-owned, for-profit utility companies with the exclusive and indefinite right to deliver power within their service territories – and reassign ownership and control of Maine’s grid to a new public non-profit entity called Pine Tree Power.  

According to the leaders of this grassroots movement, CMP and Versant peddle a deteriorating product to a captive customer base without repercussion. By placing power delivery in the hands of the state government, so the rationale goes, consumers become constituents and can better hold utilities accountable – not only to delivering low-cost, reliable power, as dictated by their franchise charter but to whatever is in the democratically-determined public interest, such as decarbonization and local resilience.

Yet the ballot measure was rejected handily. When the Associated Press called the loss, 69% percent had voted against, corroborating my anecdotal research undertaken while visiting my partner’s family in Portland during the holidays as I noticed many more signs to “Vote No” on Pine Tree Power than otherwise. Naysayers primarily argued that other avenues to pursuing climate goals and utility reform, such as adopting performance-based regulation, are substantially less risky and have a growing track record of positive outcomes.   

As it turns out (to no surprise for many), it is remarkably difficult for anyone – firm, municipality, local community, or otherwise – to extricate itself from the gravitational pull of the state-regulated monopoly utility model that reigns supreme throughout the vast majority of the United States. The reason is that Samuel Insull, mastermind of this devilishly obstinate but effective regulatory arrangement, understood exactly how to thwart the primary mechanisms of institutional change. 

According to the political economist Albert Hirschmann, there are two peaceful methods of getting what you want from an organization or firm in response to declining service quality: exercising exit and voice. Both aim to apply institutional pressure to reduce “slack”: the margin between the incumbent service provider’s best possible performance and its actual performance. Although it is impossible to actually identify and attain this theoretical upper bound of “best possible performance”, it is possible to shrink the gap between it and a subpar status quo by incentivizing better behavior on behalf of the incumbent.  

Often the calling card for advocates of free enterprise and the fruitfulness of creative destruction, “exit” refers to “exiting” a relationship with the existing institution or firm and pursuing alternative, competitive offers for provided services. By cutting off revenue, instigators leveraging exit create a sense of urgency within the incumbent by introducing an existential risk: bankruptcy. In other circumstances, using “voice” to persuade the service provider to modify behavior towards a dedicated end is preferable. Instead of expressing dissatisfaction by quietly defecting to competitive products, voice involves remaining loyal, articulating a problem, mobilizing stakeholders, and initiating change through existing institutional channels. 

We naturally group the exit option into the world of economics and voice into the world of politics yet this tendency is driven more by academic expediency than faithfulness to what occurs in practice. In his book Exit, Voice, and Loyalty, Hirschmann explains how exit works in the political arena by highlighting how voters defect to rival political parties and how immigrants escape poor governance by leaving countries of origin. He also evaluates how businesses use voice. If it is easy to switch to an alternative provider, exit is actionable and its threat is genuine. However, deep investments and “stickiness” in the existing relationship, whether via technical integrations or strong emotional bonds, increase the pain of migration and make exit ineffective, impractical, or even unthinkable. When there is no robust market of alternative offerings with reasonable switching costs, voice shines in the marketplace. 

So high switching costs, whether driven by emotional ties or an investment in the service or product offering, create loyalty, limit the feasibility of exit, increase the utility of voice, and transform the instigator’s role from customer to constituent.

Yet voice suffers from a different limitation: it takes more work. To be effective, voice requires offering a creative, captivating method of fixing the deterioration problem or generating so much discomfort within the counterparty that it is compelled to reduce slack. If the incumbent is antagonistic and/or complacent, the customer-constituent must identify the issue, articulate a solution, rally affected parties to the cause, and find points of leverage (Hirshman’s classic example of this end-to-end process is a workers’ strike). This tall order is even more difficult when the service provided is technically complex or if the institutional environment is designed to funnel advocacy through cumbersome political channels, limiting coordination and diluting bargaining power among customer-constituents. Technical complexity further diminishes the transmissibility of a new disruptive idea or “meme” by limiting the ease with which an instigator can rally support for a cause and increasing its vulnerability to the asymmetric information, lobbying, and expertise of the incumbent (which may even have control of the very information required to evaluate new ideas and create compelling narratives). 

Since voice requires more time, resources, and creativity than exit, sometimes it reverberates indefinitely and aimlessly, failing to generate substantial pressure and meaningful change to the status quo.

When institutions silence both exit and voice to maximize slack, it is known as “regulatory capture”, and this is precisely what has Mainers – and many other communities, companies, and individuals – trapped in a perpetually disadvantageous position, courtesy of old institutions that emerged during what historian Richard Hirsch calls the “utility consensus”.

After the dust settled from the Current War and centralized grids with AC power emerged as the dominant method of delivering electricity, a chaotic world of small, temporary franchises dictated which companies had the right to serve various cities and neighborhoods. Although this system created competitive pressure between electricity providers, as it gave local communities the option to “exit” status quo arrangements, it suffered from a variety of political ailments that ultimately led to its demise. First was that competitive pressure did not lower costs. Grid infrastructure, for both delivery and early thermal generation, exhibited such dramatic and obvious returns to scale that a lack of centralization left vast quantities of consumer surplus on the table. Additionally, such capital-intensive businesses struggled to raise debt at workable rates when limited franchises brought uncertainty about how long the electric service company could maintain its customer base. And perhaps most unfortunately, this decentralized, wild-west environment facilitated corruption. High-profile stories of crooked mayors and municipal representatives handing out franchises as political favors sowed distrust in local governments and generated a distaste for public power, even among Progressive politicians. 

So an unlikely consensus emerged between utility managers, Progressive politicians, academics, and eventually universities and equipment manufacturers, which proposed a new path forward paved by private, investor-owned monopolies with perpetual and exclusive franchises, regulated by technical experts from state governments – ostensibly above the fray of local politics and capable of providing unbiased technical expertise. Every component of this proposal was essential to its overall effectiveness and survival. Private, investor-owned corporations brought technical and managerial expertise alongside the lure of a pure profit motive, providing the best aspects of free enterprise; exclusive monopoly franchises ensured they could amortize massive equipment over large, captive customer bases and borrow capital at rates worthy of businesses insulated from competition, driving down costs; unbiased and supposedly apolitical state regulatory agencies consisting of technical experts would oversee investments and ensure the utilities accomplished the social objectives laid out in the regulatory charter: providing low-cost, reliable power to anyone within their service territory.

It immediately became clear that the state regulatory agencies were much less partial than intended – not only due to the whims of politicians but from intimidation tactics of the private utilities. An inconspicuous, unglamorous, and under-resourced office, many state politicians used the early public utility commissions as a dumping ground for the least capable people on the list of those owed political favors. This left them easily, and often willingly, manhandled by the utilities they were supposed to regulate, which spent much more lobbying money on shaping policy through these institutions than any third-party or public interest. As historian Richard Hirsch explains the dynamic in Power Loss:

“As early as the 1920s, commissioners had come under the influence of managers or acted in ways that pleased them, giving utility companies favorable treatment for construction plans, rate-base valuations, and rate reduction requests. Though acting in concert with power companies, commissioners still performed an important function by providing the impression that they protected customers from monopoly abuses. Therefore, it remained in the utilities’ interest to argue publicly that regulation imposed an onerous burden on them. But insiders – regulators and managers – knew better.” (45-46). 

In the first few decades of this system, none of the regulatory capture really mattered as everyone rode the economic coattails of the technological era that produced it. Increasing load growth from the proliferation of new electrical devices, the returns to scale exhibited by thermal generation technology, and cheap fuels allowed utilities to derive profits that kept investors happy while easily – almost inadvertently – providing the low-cost, reliable power required of their regulatory charter. However, several technological and geopolitical changes materialized in the 1970s that finally began threatening the fundamental assumptions underpinning this regulatory arrangement. Yet it has continued to hold under this steadily building political pressure. Why?

Samuel Insull, leader of the utility conglomerate that shaped these formative institutions, capitalized on historically unique socio-economic circumstances to bring an unfathomable level of protection to his business. Most companies work tirelessly to build and maintain a defensible moat against exit through walled-garden services, strong brands, interoperable products, or specialized data. Insull spared future generations of utility managers from this pressure by etching a moat into legislative stone.

The genius of the system Insull built is how it maximizes slack – the ability to derive profits with minimal effort – by marginalizing both exit and voice. Exclusive and indefinite franchises protect utilities from private competition on electricity delivery, forcing innovators to either target smaller markets behind the meter or build new businesses with or through the utility, nullifying exit. Meanwhile, as Hirsch has shown, public utility commissions funnel voice through inconspicuous and outmanned state regulatory agencies behind the scenes, where utilities dictate the terms of engagement. As the more concentrated, well-funded actors, utilities spend more on lobbying (with money coming directly from their adversaries, the ratepayers) within this highly technical regulatory process than any loose third-party coalition of residential and commercial customers, nongovernment entities, and academic organizations. They also hold the epistemological trump card: near-exclusive access to operational data, the source of truth for grid modeling, which puts instigators at a perpetual information disadvantage and limits their ability to make effective empirical arguments.

So Mainers are trapped. Advocates for Pine Tree power have had to summon an inordinate amount of voice to convince a disparate, disorganized population that knows little to nothing about grid modeling and electricity to care about the highly technical and arcane details of its ambitious agenda: mass municipalization as a means to attain cheaper, cleaner power. It failed (for now). Final estimates vary, but Canary reported it was outspent 35 to 1. Ratepayers neither have the institutional channels to hold utilities accountable through voice, as constituents, nor do they have the right to exit, as customers. 

However political institutions do not operate in a vacuum. Even as they remain fixed, culture and technology evolve freely around them until innovation unlocks waves of punctuated equilibrium and big bursts of regulatory change. Critically, this phenomenon tends to occur when voice and exit leverage their complementary capabilities in pockets of pressure outside the purview of the incumbent institutions. This pressure has been building due to the trends that a16z identifies in its essay.  

In the last decade, we’ve seen customers add capacity behind the meter at record rates to provide resilience, minimize interconnection delays, and avoid rising delivery costs, especially in hard-hit regions like California. Following Duncan on Twitter or listening to the DERTF podcast (the recent episode with Quincy Lee Edwards is particularly relevant) highlights no shortage of examples where entrepreneurs are helping customers build around the grid edge and circumvent the limitations of distribution utilities. 

At the same time, frustration with investor-owned utilities is escalating within local communities: towns, villages, and cities. By design, municipalization is a painful and time-consuming process, fraught with years of legal battles and expensive grid infrastructure buyouts. However, many communities now pursue a more lightweight recent innovation, community choice aggregation (CCAs), to take power supply (but importantly, not delivery) into their own hands. This is where voice has been at its best: in relatively small, tight-knit communities with unified interests when the new governing institution is not some faceless corporation but real people that constituents run into at the neighborhood grocery store.

It is within these most granular units of government where voice and exit can be freed to build real institutional pressure. Paradoxically, the best political method of realizing the full value of their complementary capabilities is akin to escaping a Chinese finger trap: relaxing the legal restrictions around the grid edge as means to incentivize experimentation and build leverage behind the incumbent utility meter.

CCAs have garnered more success than municipalization because they only seize the means of supply, not delivery, meaning the local government does not have to raise funds to buy out utility infrastructure and endure years (or decades) of legal battles. However, delivery costs have the highest growth rate out of any component of utility bills in most regions, meaning communities will eventually be forced to act. Thankfully, they can leverage an advantage that state regulators do not have: local permitting and zoning jurisdiction.

A shrewd township or municipality will avoid playing into the incumbent utility’s hands by instead cordoning off an industrial zone within its borders and commissioning an innovative developer to build new infrastructure for greenfield industrial microgrids. If executed correctly, the city avoids paying for existing grid infrastructure, streamlines fresh economic development, and brings long-lasting local job opportunities to its constituents; meanwhile, a handful of new industrial tenants share the cost of interconnection upgrades and/or onsite generation to mitigate delivery risk from the grid. After bolstering tax revenues and technical know-how from facilitating new economic activity within the industrial zones, the city can either use the windfall to fund infrastructure acquisitions or as newfound leverage to obtain better terms from incumbent IoU utilities. This is how you consume the bulk grid from the outside in.

Thus, the two great macro-level energy trends that have emerged in response to dissatisfaction with IoU service – Community Choice Aggregation carrying the banner of voice and DER defection carrying the banner of exit – are unified outside the scope of existing institutions. But as evident by the doomed request from Sunnova to the CPUC to make a micro-utility in California, playing the game through existing state institutional channels will kill any hope of building leverage. It must be clandestine. Only local governments have both the authority and incentive to relax the finger trap and revoke, or at least reclaim power over the administration of, franchise rights. They have the authority in that they control zoning and permitting jurisdiction over their territory and the incentive in that they are publicly accountable to small, motivated populations capable of effectively using voice, whose primary interests are some combination of decarbonization and economic development.

With local communities aggressively courting economic development and private companies competing furiously to win the operating rights of these industrial zones, the “Swiss Cantons” of the grid will be either fully public or publicly permissioned private networks operating side by side. Hindered by high delivery costs, lack of reliable power, and interconnection delays, industrial consumers – a formidable political coalition long in cahoots with the IoUs – will seek alternative institutional arrangements allowing them to thrive in this wave of onshoring and growth in AI computing. Any new offering with low aggregate delivery costs, access to reliable power, and fast interconnection timelines will win their business. Providing a new private micro-utility with a blank slate to construct an innovative service offering aimed at servicing these customers will allow entrepreneurs to figure out exactly what combination of software, batteries, distributed generation, advanced monitoring hardware, and other modern marvels of grid tech – marvels that investor-owned utilities keep pinned to the x-axis of the adoption curve – can create highest-quality service at the lowest price.

Early adopter communities will inevitably encounter fierce resistance from neighboring IoUs and state regulators but the floodgates will open once a city allows and encourages an industrial microgrid developer to violate franchise rights and then jointly countersue the utility. Civil disobedience on behalf of a few pioneer cities, villages, or townships with the right combination of moral standing and commercial interests will ask for forgiveness, not permission, and usher in a new era of bottom-up grid development – one that may resemble the pre-Insull world of one hundred years ago, but with different technological assumptions about what institutional design can succeed in the digital era.

Thank you to Jed Trott and Elias Hatem for reviewing earlier versions of this essay.