The biggest barrier to the clean energy transition isn't technology — it's time. According to Stanford economist Hunt Alcott, the electric power industry moves more slowly than almost any other sector because the infrastructure we build today will remain in the ground for 50 to 100 years. Combined with political gridlock and the growing energy hunger of AI-powered data centers, getting to a clean grid is far more complicated than simply building more solar panels and wind turbines. Here's what the latest research from Stanford's leading environmental economists reveals about where we stand — and where we're headed.
What Is the Biggest Barrier to the Clean Energy Transition?
When Professor Alcott first worked in the energy industry back in 2002, the power plants he modeled were already 40 years old. Today, most of those same plants are still generating electricity. That's not a coincidence — it's a defining feature of the industry.
04:12
Professor Alcott explains why the electric power industry changes more slowly than any other sector
Watch at 04:12 →
Unlike consumer tech or software, energy infrastructure requires enormous upfront capital investment with a payback horizon measured in decades. A natural gas plant built today will still be running in 2075. This means the decisions policymakers and utilities make right now will shape the carbon intensity of the grid long after most of us have stopped thinking about them.
Beyond infrastructure inertia, Alcott points to a second major obstacle: political constraints. Reaching consensus on climate policy in the United States has been slow and contentious. Without a stable, long-term policy signal, investors hesitate to commit capital to new clean energy projects — and the transition stalls.
- Infrastructure longevity: Power plants last 50–100 years, locking in today's choices for generations.
- Political gridlock: The US has struggled to build lasting consensus on carbon pricing or clean energy mandates.
- Historical cost of renewables: Wind and solar were more expensive for decades, though that has now changed dramatically.
How Is AI Changing Energy Demand and the Power Grid?
Just as electricity demand had flattened out, artificial intelligence arrived and changed the math entirely. Data centers powering large language models and cloud computing are consuming electricity at a scale that is reshaping utility forecasts across the country.
For years, electricity demand grew at roughly 1–2% per year before plateauing. Now, thanks largely to AI infrastructure buildout, demand is accelerating again. This creates a dual challenge: we need more power, and we need that power to be cleaner.
08:45
The discussion on AI-driven data center electricity demand and its effect on grid planning
Watch at 08:45 →
But AI isn't just a source of new demand — it also holds potential as a tool for optimizing the grid itself. Alcott highlights two specific opportunities:
Smarter Transmission
Many existing transmission lines are underutilized relative to their actual technical capacity. AI-driven optimization could allow operators to push more power through the lines we already have, reducing the need to build new natural gas plants just to meet peak demand.
Demand Response
When electricity is scarce — a summer afternoon when air conditioners are running full blast and everyone is charging their EVs simultaneously — the grid strains under peak load. AI can help manage demand response programs that nudge consumers to shift usage, flattening those dangerous spikes without requiring new generation capacity.
So while AI increases the total amount of electricity the world needs, it also offers tools to use that electricity far more efficiently. Whether it becomes a net positive or negative for sustainability depends on how quickly the grid can decarbonize to meet the new demand.
02:30
Alcott cites Janet Currie's research showing Clean Air Act policies helped disadvantaged communities most
Watch at 02:30 →
Can Environmental Policy Be Both Efficient and Equitable?
One of the most persistent debates in environmental economics is whether policies can be both economically efficient — delivering maximum environmental benefit per dollar spent — and equitable, protecting the most vulnerable communities from bearing the worst burdens of pollution.
The conventional wisdom says there's a tradeoff. Alcott's answer is more nuanced: sometimes efficiency and equity are in tension, but sometimes they point in exactly the same direction.
He cites research by economist Janet Currie and her co-authors, which examined improvements in fine particulate matter (PM2.5) air quality across the United States over the past two decades. The main driver? The Clean Air Act's nonattainment standards, which required the dirtiest counties — those exceeding legal air pollution limits — to clean up fastest.
Here's the equity insight: the dirtiest counties also happened to be home to the most disadvantaged populations. By targeting the worst pollution first, which is the efficient thing to do given how harmful fine particulates are, the policy simultaneously delivered the greatest health benefits to the people who needed them most.
18:20
Stanford researchers break down the equity and trade provisions of IRA electric vehicle tax credits
Watch at 18:20 →
This doesn't mean the tension between efficiency and equity never exists — it does. But the research suggests that as countries demand higher environmental quality over time, the benefits of cleaning up pollution will disproportionately reach lower-income communities that have historically lived with the worst air and water.
How Do Electric Vehicle Tax Credits Actually Work?
The Inflation Reduction Act (IRA) introduced significant changes to electric vehicle tax credits, and Alcott's research group at Stanford is actively studying the effects. Two features of the new credits stand out as particularly novel.
Trade Restrictions and Onshoring
For a buyer to qualify for the full EV tax credit under the IRA, the vehicle must meet strict requirements about where its battery components are manufactured. This is a deliberate attempt to onshore EV and battery production to the United States, reducing dependence on foreign supply chains — particularly Chinese ones.
21:05
The debate over Chinese EV tariffs and whether US manufacturers qualify as infant industries
Watch at 21:05 →
Equity Provisions
Previous EV subsidies had a well-documented equity problem: new electric vehicles are expensive, so the subsidies mostly flowed to wealthy buyers. The IRA attempted to fix this by introducing credits for used electric vehicles, along with income caps and vehicle price limits to ensure the benefits reach middle- and lower-income households.
Alcott's group is publishing research evaluating both the trade dimensions and the equity outcomes of these credits — work that will be critical as policymakers debate whether to extend, expand, or modify EV incentives going forward.
Should the US Allow Chinese Electric Vehicles to Be Sold?
Here's a striking fact: in the history of the modern United States, there is essentially one consumer product that Americans are effectively barred from buying due to prohibitively high tariffs. That product is the Chinese electric vehicle.
Alcott frames this as a genuinely open question with legitimate arguments on both sides.
The case for allowing Chinese EVs: Greater competition could accelerate the energy transition by putting affordable, high-quality electric vehicles in more American driveways faster. It would also pressure domestic manufacturers to innovate more aggressively — a dynamic well-understood in industrial economics, where competitive pressure drives productivity gains through both the adaptation of existing firms and the creative destruction of weaker ones.
The case against: Protecting a domestic EV industry in its early stages — treating it as an infant industry — may be justified if it allows American manufacturers to reach the scale and efficiency needed to compete globally over the medium term. History offers mixed evidence here. Protectionism in the Soviet bloc and parts of Latin America often produced inefficient, uncompetitive industries. But targeted industrial policy in parts of Asia has sometimes successfully nurtured globally competitive sectors.
Professor Barnett draws an important distinction: protecting a mature incumbent like General Motors is very different from nurturing a genuinely nascent American EV industry. The policy question of whether today's US EV makers qualify as infant industries worthy of protection is, as Alcott acknowledges, genuinely unresolved.
What Is the Social Cost of Carbon and Why Does It Matter?
Before policymakers can design effective climate policy, they need to answer a deceptively difficult question: how much does one additional ton of carbon dioxide actually cost society? This is what economists call the social cost of carbon, and it sits at the heart of environmental economics.
Getting this number right matters enormously. Set it too low and you'll underinvest in emissions reductions, leaving climate damage on the table. Set it too high and you'll impose costs on the economy that exceed the benefits. Alcott identifies the social cost of carbon as perhaps the most important output of the non-market valuation branch of environmental economics — the field dedicated to putting dollar values on things that don't have market prices, like clean air, a stable climate, or a healthy ecosystem.
How Can Better Transmission Lines Speed Up Decarbonization?
Wind blows in Wyoming. The sun shines in the Mojave Desert. But the people who need that electricity live in Chicago, Los Angeles, and New York. Bridging that geographic gap is one of the central infrastructure challenges of the energy transition.
The obvious solution — build more transmission lines — is correct but incomplete. Alcott's research group is focused on a less-discussed opportunity: the transmission capacity we already have is significantly underutilized. Through better market design and technology-enabled optimization, we could move substantially more clean power across existing lines without laying a single mile of new cable.
Combined with demand response programs that smooth out consumption peaks, smarter use of existing transmission infrastructure could meaningfully reduce the number of new fossil fuel plants that utilities would otherwise need to build to meet rising demand — including demand driven by AI.
The clean energy transition is not a single problem with a single solution. It's a layered challenge involving aging infrastructure, political economy, market design, equity, and now the wild card of artificial intelligence. What Stanford's environmental economists make clear is that the tools to solve it — better policy, smarter markets, and rigorous research — already exist. The question is whether we move fast enough to use them.
