Electrification? We Are Already On The Way
In 1960 only 2% of U.S. homes were heated with electricity. Today it’s 38%.
U.S. households burn vast amounts of fossil fuels on-site for home heating: 2.7 trillion cubic feet of natural gas, 2.9 billion gallons of fuel oil, and 2.5 billion gallons of propane. In terms of carbon dioxide emissions, this is equivalent to having 40 million cars on the road.
Looking at all this fossil fuel consumption, policymakers and climate advocates are eyeing electrification of home heating as a potentially low-cost path to reduce carbon dioxide emissions.
But electrification of home heating is nothing new. In today’s blog, I look at data documenting a large increase in electrification of U.S. home heating over the last 60 years.
While well short of what the United States would need for “deep decarbonization”, this trend suggests that getting people to switch to electricity may not be as hard, or as expensive, as previously believed.
The Rise of Electric Heating
Only 2% of U.S. households in the 1960 Census reported using electricity as their primary heating fuel. Electric heating has increased steadily since that time, reaching 18% in 1980, 29% in 2000, and 38% in 2017. This is a large increase; much bigger than I would have guessed, frankly.
The geographic pattern is interesting. Perhaps most strikingly, the maps show how electricity has grown to become the dominant form of heating in the Southeast — 50%+ throughout the region and 90%+ in Florida. Electric heating is also prevalent throughout the West and Midwest, albeit less than 50% in most states.
Electricity Prices Matter
Why has electricity’s market share grown? I have not seen much written on this topic, but I think part of the explanation is electricity prices.
In 1960, U.S. residential electricity prices were 40% higher in real terms than they are today. Prices dropped sharply during the 1960s and have remained approximately flat during this long period of growth in electric heating. We would expect there to be a negative correlation between electricity prices and electric heating, and this appears to be the case.
New home construction matters too. During this period of low electricity prices, there have been relatively more new homes built in the sunbelt. It’s the interaction between electricity prices and new home construction that drives these trends.
States with cheap electricity tend to have more electric heating. In 1960, the four states with the most electric heating were Washington, Oregon, Nevada, and Tennessee. This is no coincidence. These four states had access to cheap Federal electricity — via the Bonneville Power Administration and the Tennessee Valley Authority.
This correlation continues to date. The Southeast and Northwest have lower than average electricity prices — and more electric heating. California has higher electricity prices – and less electric heating than neighboring states.
Clean and Convenient
Another potential explanation for the growing use of electric heating is household income. Median household income approximately doubled in the United States between 1960 and today. As incomes have increased, U.S. households have increasingly selected electric heating.
I think this makes sense. Electric heating is cleaner and more convenient than other forms of heating. There is no on-site combustion, so there are no on-site emissions. Also with electric heating you do not have to deal with a furnace, storage tank, or ductwork. You don’t have to schedule deliveries like you do with heating oil.
With electric heating, it is also particularly easy to control the heat levels in different rooms. Many new multi-unit buildings use electricity, simply because it so easy to work with and not capital-intensive.
There are certainly other factors that matter too. For example, the overall “scale” of heating demand matters a lot for this decision. Electric heating has high incremental costs, so it makes most sense in relatively mild climates.
Households in cold climates tend to choose natural gas because of the low price per unit of heating. But not everyone has access to natural gas. In many rural areas with low population density there are no natural gas distribution lines. This helps explain the popularity of electricity in North Dakota and South Dakota, for example.
What does this mean for electrification policy?
As my colleague Max Auffhammer recently wrote about, the city of Berkeley has become the first U.S. city to ban natural gas in new homes. This evidence on historical U.S. heating choices has a couple of implications for policies like this.
First, while there has been steady, continual progression towards electric heating, the pace of change has been slow. The building stock turns over slowly, and there is considerable inertia with home heating choices. Thus, whatever policy interventions are made in this sector, policymakers will need to be patient.
Second, these data imply that households like to heat their homes with electricity. Before I pulled these data, I had not realized the degree to which electricity’s market share has grown over time. Particularly in relatively warmer climates, U.S. households have a strong revealed preference for the cleanliness and convenience of electricity.
Bottom line: if electrification is the road to GHG reduction, it may not be as steep a road as some people think. The key will be avoiding prices that over-penalize electricity use.
Keep up with Energy Institute blogs, research, and events on Twitter @energyathaas.
Suggested citation: Davis, Lucas. “Electrification? We Are Already On The Way” Energy Institute Blog, UC Berkeley, November 4, 2019, https://energyathaas.wordpress.com/2019/11/04/electrification-we-are-already-on-the-way/
Lucas Davis View All
Lucas Davis is the Jeffrey A. Jacobs Distinguished Professor in Business and Technology at the Haas School of Business at the University of California, Berkeley. He is a Faculty Affiliate at the Energy Institute at Haas, a coeditor at the American Economic Journal: Economic Policy, and a Research Associate at the National Bureau of Economic Research. He received a BA from Amherst College and a PhD in Economics from the University of Wisconsin. His research focuses on energy and environmental markets, and in particular, on electricity and natural gas regulation, pricing in competitive and non-competitive markets, and the economic and business impacts of environmental policy.
Heat pumps are efficient, but they bring with them inescapably electric resistance heating for water heaters, clothes dryers, and cooking. In the coal/gas dominated South, they are all worse choices than burning gas (though in a place like California, an electric water heater with controls and a strong RPS can actually be fairly efficient). Electric cooking is a just plain nasty load for the grid to accept. People want to eat dinner when they want to eat dinner, so electric stoves and ovens end up adding to peak loads from say 500 to 730 pm, and adds to ramping requirements in the winter season in places like California where there is lots of solar energy. So it’s a load that’s coincident with peak (the headline peak for the remaining winter peaking utilities and hours close to peak – as the peak marches later – for summer peaking utilities) and has a very low load factor (about 20-25% against a winter peak from a calculation I made relating to Ontario Hydro a couple of decades ago).
While what you say is true for most currently installed technologies, it’s not so true for new technologies. Heat pumps are being used particularly for water heaters, and dryers are now available that are heat pumps as well. These are now approaching cost effectiveness, especially where electricity is cheap. Also induction cooktops and convection ovens are more efficient and cook even more quickly than gas.
Given the shift in peak load from afternoon to evening, the result has been a net reduction in the absolute peak load. The question is whether new resources need to be added to meet that peak load. The load shift may mean that the net reduction is less than it would be otherwise, but that’s a different issue than pure load building. That the CAISO peak load has remained largely flat or even declined for a decade illustrates this effect.
Which brings us to the incremental resources used to meet this electrification. Do we see that the added load will be met by increased coal output, or added renewables such as solar and wind? The current evidence with coal retirements appears to be the latter. So the added emissions are minimal. (Poorly run production cost models fail to capture this effect.)
When my children’s elementary school wanted to do ethernet wiring I advised against it. Because i knew wifi was coming.
New building electrification us fine, but converting from gas heating ie ‘electrification’ is going to be expensive retrofit. Who knows we might get high power microwave transmission [as in James Bond ‘Golden Eye’] in the next 30-50 years, and the retrofit would have been a waste.
Where electrification is happening is precisely where coal is king of off-peak power and there are few renewables. It has been pushed by lower winter electric rates and declining block rates – with outright lower rates for space heating in some states, as well as gas company insensitivity to revenue requirements making them less competitive. So electrification has been a minus for improving our GHG steps. I have been pushing for less promotional winter residential rates where I have worked over the years (Arkansas, Texas outside of ERCOT, etc.).
Jonathan, good point. Like “renewable” electricity, heat pumps are great if consumers live in the right areas – i.e., they don’t need much heating at all. If they live in colder climes, I guess carbon emissions are their own fault.
“Electric heating is cleaner and more convenient than other forms of heating.”
Lucas, what’s your basis for these conclusions? Whether it’s cleaner than gas depends on many factors, including:
1) Source(s) used to generate electricity, and their availability when needed
2) Line losses of local grid
3) Whether “clean” includes CO2, or only NOX/SO2/particulates
4) Whether consumers consider fast heating a convenience (if so, gas wins hands-down)
5) Local climate
Electric heating is the clear emissions winner, in California, only if we’re wise enough to end the manic push by our state’s gas and renewables industries to shut down Diablo Canyon Power Plant. France and Germany provide textbook examples of why replacing carbon-free nuclear with renewables + gas is a loser under all circumstances – especially, if we envision a future of electric transportation and desalination.
Thanks for this clarification, Carl. In the next sentence I say, “There is no on-site combustion, so there are no on-site emissions.” So I was referring to electric heating being cleaner in terms of on-site emissions. In other blog posts we’ve written frequently about total system emissions, e.g. in this recent post we write about how reduced emissions intensity of the U.S. electric sector mean that grid-powered EVs are cleaner than gasoline vehicles throughout the western United States and in virtually all cities of the eastern United States.
In looking at electrification, it’s also important to look at the truly incremental resources added to meet the incremental load. In most of the U.S. now, most of those incremental resources are renewables. The share of renewables has been increasing everywhere as they become cost competitive on a long run basis. The error is in focusing on the hourly markets where gas is the “marginal” (versus incremental) resource. Short run marginal resources are not particularly useful at emission benchmarks when the incremental resources have no short run marginal costs, so they don’t show up in the hourly marginal markets.
Richard, there is no “incremental load” on an electricity grid. Load is a continuous function of the demand of millions of electricity consumers over time – it doesn’t increase or decrease incrementally, but in a smooth curve. Similarly, considering renewables an “incremental resource” is a backhanded way of charging customers for intermittent electricity generation averaged over a period of time, as if customers are receiving “x megawatts for y hours.”
You know, and I know, it doesn’t work that way. There’s no assurance, in advance, any renewable source of electricity will be able match demand. Ever. Given that unreliability and variability, natural gas must be standing by to make renewables (particularly wind and solar) useful at all, and I have yet to see why any source that dependent on fossil fuel should have a place in a clean energy future.
New generating resources are not added kWh by kWh to the grid. Even your beloved nukes come in 1,000 MW lumps. Electricity generally comes from new large capital intensive projects (although rooftop DER is turning that notion on its head), and those added projects, not the minute by minute marginal changes, are what are used to meet those added loads. Even for rooftop solar, residents are looking to displace utility supply in its entirely or to add a new electric vehicle. All of these are incremental additions and those increments are the key to thinking about how emissions will change with electrification. You can fall into the mythology of marginalism as your means of looking at the world, but it’s not a valid view.
Richard, load is demand, not generation, and neither is measured “incrementally” in electricity transmission. Both are continuously variable, depending not only on what is consuming electricity but how much it’s consuming at any given moment.
Neither the phrase “minute-to-minute marginal changes,” nor measuiring generating resources in units of energy (kWh) make any sense at all in an electrical context.
When your aren’t familiar with basic principles of electricity what units are used it’s difficult to understand the point you’re trying to make.
You’re speaking nonsense. Transmission lines are built in increments of 100s to 1000s of MWs. And of course generation is measured in kWhs. The addition of kWs of generation capacity translates into kWhs of output. I am pointing out when a policy adds thousands of GWH of new load due to a new policy such as electrification, that load must be met with thousands of GWH of new generation resources (not incremental output from existing resources.) That means the appropriate metric for additional emissions is from those new resources, not from existing resources.
Nice try Richard, to confuse “generation resources” with “generation” retroactively. Of course generation can be measured in kWh. Solar and wind farms (i.e., generation resources), however, are not measured in kilowatthours – at least by anyone with a fundamental understanding of energy.
What’s a “minute-to-minute marginal change”? I missed an explanation of that one, maybe you jotted it in a margin somewhere. And you’re already back to measuring resources in terms of energy, with “thousands of gigawatthours of new energy resources.” Were you imagining a mutli-gigawatthour solar farm, perhaps?
I will accept your criticism of speaking nonsense as a compliment, given it comes from an expert.
I have no idea where you are trying to head with your argument. Are you trying to say that large scale electrification of buildings and transportation will result in increased output from existing gas generation despite 1) the increasing RPS mandate that requires increasing shares of renewables for existing load plus for new load 2) the capacity limit of the existing gas fleet that restricts any additional gas output during peak hours, 3) that gas generation in California has dropped precipitously since 2008, and 4) the procurement plans by California’s LSEs to meet all new load with renewables? Note that we still don’t have commercially available truly dispatchable nuclear power, so your solution would still be nuclear + gas.
Very interesting piece. I too was surprised at the extent of growth in electric heating. Could one factor also be the rise in household air conditioning? If an owner decides that they want air conditioning, it is a pretty easy next step to get electric heat as part of the package (by installing a heat pump). I know that when our furnace went out in Richmond CA, the HVAC contractor certainly pushed hard for a heat pump, which had the added benefit of providing A/C.
Dr. Davis – you are ignoring a bit the technology factor. Many TVA and BPA customers don’t have access to Natural Gas, only electricity and propane and propane is pretty costly. Heat pumps are the technology that has become readily available since 1960 and it both heats and cools, hence the adoption in the SE where cooling dominates. BPA service territory in the Pacific NW was dominated by electric resistance heat until the energy efficiency efforts began in the early 1980s when heat pumps were just becoming available. Cooling has less value in the NW though that is changing. Economics and the interaction with technology as well as climate. Climate change has certainly made the need for cooling and heating greater in the SE since the 1960s as well.
In comparing electric penetration in the North to that in the Southeast you mention incremental costs. No expert, but I have heard that as an issue for electric resistance heating but not heat pumps; and that the disadvantage of heat pumps is that they are not very effective in winter in cold climates. If the greater penetration of electric heating in the southeast is heat pumps, that would explain it.