
数据中心的争论在全国范围内持续激化,争议往往集中在一个问题上:数据中心在多大程度上推高了电费?
The data center debate rages nationwide, and it often centers on one question: How much are data centers driving up electric bills?
但大多数美国人可能会惊讶地发现,即使未来没有任何新的超大规模数据中心出现,他们的电费也会继续上涨。这引发了一个更具揭示性的问题:为什么美国仍在使用一个为20世纪初建立的、已有百年历史的公用事业模式来运行其经济?
But most Americans would be surprised to learn that their bills will keep climbing—even if no new hyperscaler data centers ever show up. And that raises an even more revealing question: Why is the U.S. still running its economy on a 100-year-old utility model built for the early 1900s?
在现代使用这种过时的电力输送系统,会产生深远而广泛的后果。过去五年,电费上涨了30%,典型的美国家庭每年多支付300美元。至少有八个州的公用事业公司正在提前切断电力供应。一些州正在应对由其自身公用事业设备引发的灾难性野火威胁。与此同时,公用事业公司为满足预测中的超大规模数据中心需求而采取的过时做法,威胁着给普通电费缴纳者带来更高的成本。
Using this antiquated electricity delivery system in the modern era has large and widespread consequences. Electricity rates have risen 30% in the past five years, and the typical American household pays an extra $300 annually. Utilities in at least eight states are preemptively shutting off electricity service. And some states are responding to threats of catastrophic wildfires often sparked by their own utility equipment. Meanwhile, obsolete utility approaches to meet forecast hyperscaler data center demand threaten even higher costs for ordinary ratepayers.
作为一个国家,我们应该承诺采用并连接太阳能和电池储能等新技术,让个人和社区对电力生产和供应拥有更多控制权,并提供满足电力需求的最便宜、最清洁的方式。然而,旧的垄断公用事业体系为公司创造了强大的动机,使其违背公众对清洁、廉价能源的需求,并阻碍现有的公共监督机制。加速本地电力生产和韧性的技术改进,进一步扩大了过时的公用事业模式与现代电力系统所带来的机遇之间的差距。
As a country, we should commit to adopting and linking new technologies like solar and battery storage, giving individuals and communities more control over power production and supply, and offering the cheapest and cleanest way to meet our power needs. However, the old monopoly utility system creates powerful incentives for companies to act against the public need for clean, cheap energy and to thwart existing public oversight mechanisms. Accelerating technological improvements for local power generation and resilience further widens the gap between the outdated utility model and the opportunities of a modern electricity system.
公平地说,我们那套陈旧的体系在当时是有道理的。自从发电业出现以来,政治领导人就相信这种基本资源的公共效益,并认为垄断供电——每个家庭或企业只有一套电线——是审慎的做法。随后,各州一个接一个地达成协议,允许营利性公司垄断电力供应,以换取州政府的监管。这种逐州的监管模式很快就力不从心了,因为公用事业公司在未受监管的行业投机,并隐瞒成本。
To be fair, our archaic system made sense at the time. Since the advent of electricity generation, political leaders believed in the public benefits of this essential resource and that a delivery monopoly—a single set of wires to each home or business—was prudent. Then, one after another, states struck bargains to grant for-profit companies a monopoly over electricity delivery in exchange for state oversight. This state-by-state regulatory model quickly fell short, as utilities speculated in unregulated industries and hid the costs.
20世纪20年代和30年代大型公用事业控股公司的倒闭助长了大萧条,迫使制定新的联邦规则。1935年,《公用事业控股公司法》做出了几项旨在遏制公用事业行业过度行为的改变,包括拆分多元公用事业控股公司,并制定准则,防止威胁公共利益的合并再次发生。
The collapse of large utility holding companies in the 1920s and 1930s contributed to the Great Depression, demanding new federal rules. In 1935, the Public Utilities Holding Company Act made several changes meant to curb utility industry excesses, including breaking up multi-utility holding companies and setting guidelines to prevent a recurrence of mergers that threatened the public interest.
这一修订后的监管框架又维持了50年,但随着“规模越大越好”的经济理念在20世纪80年代和90年代的美国政治中占据主导地位,市场整合的热潮也席卷了公用事业行业。从20世纪90年代中期到2015年,合并使投资者拥有的公用事业公司数量减少了近一半,而在此期间用电量却上升了20%。市场竞争的丧钟于2005年敲响,当时国会废除了1935年立法中的关键条款,取消了对公用事业公司政治捐款的禁令,大幅削减了联邦监管,并显著增加了政治影响力。
This revised regulatory framework held for 50 more years, but as an economic philosophy of “bigger is better” took hold in our national politics in the 80s and 90s, the fervor of market consolidation swept up the utility industry, too. From the mid-90s to 2015, mergers shrank the number of investor-owned utilities by almost half, even as electricity use rose 20%. The death knell of market competition came in 2005, when Congress repealed key provisions of its 1935 legislation and lifted a ban on utility political contributions, sharply reducing federal oversight and dramatically increasing political influence.
公用事业公司越来越少、规模越来越大,这意味着股东拥有更多权力,而消费者得到的服务却更差。
Fewer and larger utility companies have meant more power for shareholders and poorer service for consumers.
公用事业公司不仅通过扼杀潜在竞争对手来巩固这种权力,还利用基础设施规划规则中的漏洞,建设自身的宠项目,从而有效排斥新进入者。它们游说各州制定“优先购买权”法律,以便优先拥有新的电线杆和电缆。它们还为新的太阳能阵列和电池的电网接入流程增加了层层繁琐耗时的审查和高额费用。此外,它们还积极游说削减消费者从屋顶太阳能项目中获得的节能收益。
Utilities have only reinforced that power by crushing potential competitors. They’ve used loopholes in infrastructure planning rules to build pet projects that effectively excluded new entrants. They’ve lobbied states for "right of first refusal" laws to get first dibs on owning new poles and wires. They’ve added layers of detailed and time-consuming review and high fees to the grid connection process for new solar arrays and batteries. They’ve also aggressively lobbied to reduce the energy savings consumers earn from rooftop solar projects.
这种陈旧且昂贵的公用事业模式,与当前电力生产和分销领域正在进行的显著技术创新越来越格格不入。Rewiring America的研究表明,数据中心预计的几乎全部需求,都可以通过住宅或商业清洁能源资产(如太阳能、电池和热泵)来满足,其速度比公用事业偏好的大型发电厂和输电线路等投资更快、成本更低。同样,社区太阳能项目只需几个月即可建成,并可在许多“订阅者”(包括租户以及通常被排除在屋顶太阳能采用之外的其他人)之间共享电力。
This old and expensive utility model is increasingly at odds with remarkable technological innovations underway in the production and distribution of electricity. Research by Rewiring America suggests that nearly all the forecast demand from data centers could be met with residential or business clean energy assets such as solar, batteries, and heat pumps, much faster and more affordably than utility-preferred investments like large power plants and transmission lines. Similarly, community solar projects can be built in mere months and share electricity among many “subscribers” (including renters and others typically left out of rooftop solar adoption).
这就是为什么像我这样的专家认为,本地资源可以直接降低数百万消费者的能源成本,同时使电网在面对野火或风暴等气候事件时更具韧性。
This is why experts like myself believe that local resources can directly lower energy costs for millions of consumers and simultaneously make grids more resilient against climate events such as wildfires or storms.
例如,加州家庭和企业主安装的数吉瓦屋顶太阳能,仅在2024年就为全州所有公用事业客户节省了超过20亿美元。许多加州人已将电池连接到其太阳能系统,使其能够在批发价格飙升期间向州电网调度电力,从而降低所有人的成本。但该州的三大投资者拥有的公用事业公司只有抱怨,因为这些来自消费者投资的消费者节省并未带来巨大的股东利润。
The gigawatts of rooftop solar installed by California home and business owners, for example, saved all in-state utility customers over $2 billion in 2024 alone. Many Californians have attached batteries to their solar systems, allowing them to dispatch power into the state’s grid in periods of sky-high wholesale prices, lowering costs for everyone. But the state’s big three investor-owned utilities have only complaints, because those consumer savings from consumer investments don’t result in big shareholder profits.
我们可以防止数据中心——或下一个巨大的能源需求——推高能源账单,但这需要承认两个基本事实。
We can prevent data centers—or the next big energy demand—from spiking energy bills, but it requires acknowledging two fundamental truths.
首先,当今电网的高昂弊端在于过于宽泛的垄断权力,这种权力让以营利为目的的电网运营商与能够为自己和他人生产和管理清洁电力的创业者和客户相对立。例如,公用事业公司之所以能阻挠屋顶太阳能,是因为它们拥有公共授予的电线杆和电线垄断权,太阳能阵列需要连接这些设施。
First, the costly failures of today’s grid lie in the overbroad monopoly powers that pit for-profit grid operators against the entrepreneurs and customers who can generate and manage clean electricity for themselves and others. Utilities can only block rooftop solar, for example, because they have a publicly granted monopoly over the poles and wires to which the arrays connect.
其次,只有拆分以营利为目的的公用事业公司,并从电力输送中剔除利润动机,才能终结那些驱动以营利为目的的公用事业公司滥用其市场和政治权力的扭曲激励。
Second, only breaking up for-profit utilities and removing the profit motive from electricity delivery can end the perverse incentives that drive for-profit utilities to abuse their market and political power.
如果将本地电力分配作为一项公共服务,重组后的公用事业公司只有在提供有价值的服务时才能获利,而不仅仅是建造某样东西。它们为人工智能数据中心(或其他任何事物)提出的电力解决方案,应当证明其成本效益与独立生产商提供的能源,或客户网络部署的本地太阳能、电池和热泵所提供的能源一样具有成本效益。
If we make local electricity distribution a public service, restructured utilities will profit only when they provide a valuable service, not just build a thing. Their proposed power solutions for AI data centers (or anything else) should have to prove as cost-effective as the energy provided by independent producers or networks of customers deploying local solar, batteries, and heat pumps.
就像联邦政府五十年前拆分AT&T一样,将运营网络的公共职能与提供服务的竞争职能(AT&T案例中的电话和调制解调器,公用事业案例中的本地太阳能电力)分开,可以释放更低的成本和创新——例如催生现代互联网的原始光纤网络。
Like the federal government’s breakup of AT&T fifty years ago, separating the public function of running the network from the competitive function of supplying services (telephones and modems in the AT&T case, local solar electricity in the utility case) can unleash lower costs and innovations –– such as the original fiber optic network that birthed the modern Internet.
电力垄断模式已经失败,无法将中央计划与私营利润与现代技术所实现的电力民主相调和。如果我们打破阻碍的垄断公司,就可以重组公用事业体系,利用由负责任的电网运营商运营的公平开放市场中的最佳资源。让公众重新掌控,将使我们能够建设一个更清洁、更高效、更具韧性的电网,更好地应对现代经济的挑战,并兑现这项美国人可以依赖的公共服务的承诺。
The electricity monopoly model has failed, unable to reconcile central planning and private profits with electricity democracy enabled by modern technology. We can restructure our utility system to tap the best resources from fair and open markets run by accountable grid operators, if we break up the monopoly corporations that stand in the way. Putting the public back in charge would allow us to build a cleaner, more efficient, and resilient grid, better able to meet the challenges of the modern economy, and make good on the promise of this public service that Americans can depend on.