Robust Access to Electricity in America's Hottest Cities
Redundancies and Contingency Plans in an Increasingly Uncertain World
This is a HOT summer and many Americans are cranking up their air conditioners and people in previously cool areas (think San Fran and Seattle) are installing AC. Nobody gains direct utility from an air conditioner. AC users seek to use air conditioning to offset the heat but they need electricity to operate the AC.
While the United States has a highly reliable electricity grid, blackouts do occur. Such blackouts on a hot day in Phoenix could cause death and discomfort. How would hospitals function? One recent clickbait article claims that 1/2 of the population would suffer greatly on such a day. This is an important claim. How do we adapt?
We released an NBER paper on this topic last year.
U.S Electric Utility Adaptation to Natural Disasters Shocks and Green Power Mandates
Robert Huang & Matthew E. Kahn
Access to electricity is a crucial determinant of quality of life and productivity. The United States has a highly reliable electricity grid, but it faces new resilience challenges posed by more intense natural disasters and ambitious green power requirements. Using a US electric utility panel dataset from 2013 to 2020, we document that natural disasters disrupt service, but utilities have made progress in adapting to such shocks. Over the last decade, utilities have faced a tradeoff between achieving local carbon mitigation goals and offering reliable power access. We discuss alternative approaches to attenuate this tradeoff.
My co-author is a USC undergraduate! I am an educator!!
What I want to discuss here is the simple economics of robustness.
“Robustness” refers to situations where decision makers know that they do not know the probability of an event.
Case #1; I flip a coin. Here you know the probability that I flip “heads” = 50% and the probability I flip tails = 50%.
Suppose that God flips a coin to decide where there will be blackout in Phoenix. If a Blackout occurs, there are 5 million people in the area who each would lose $2,000 in the money equivalent of the pain they will suffer to live the day without air conditioning and having access to electricity.
Given these assumptions, the expected loss each day from power blackouts in Phoenix equals; .5*5 million * 2000 = 5 billion dollars in damage a day. That’s huge!!
Each individual suffers an expected loss each day of $1000 = .5*2000. Such individuals would be willing to pay up to $1000 a day for a technology or a service that would guarantee them access to power. This would create a huge market for entrepreneurs to innovate and provide such “diesel backup power supplies”. Amazon already sells many of these. At any point in time, richer people will be more likely to buy these self-protect technologies and this means that the poor face greater blackout risk because they will be less likely to invest in a contingency plan.
Our 2017 innovation paper optimistically argues that innovators efforts and their competition leads to higher quality (cheaper) adaptation products over time that eventually can be purchased by the poor.
NOTE that the diesel power generator does not run on electricity. This means that the owner of this product can access electricity even if the grid is down. For an owner of a generator; his probability of not having access to electricity equals;
Prob(no power) = prob(grid broken)*prob(diesel generator doesn’t work)
In this example the prob(grid broken) = .5
suppose that the prob(diesel generator doesn’t work) = .05
then the probability a person who owns a generator doesn’t have power = .5*.05 = .025.
Now suppose this guy buys a second generator.
His new probability of not having power =
Prob(no power) = prob(grid broken)*prob(diesel generator #1 doesn’t work)*prob(diesel generator #2 doesn’t work) = .5*.05*.05 = tiny.
My Point?
Rational people who seek to adapt to high heat and who are risk averse have strong incentives to protect themselves against the failure of major infrastructure (such as the power grid) by investing in backup plans. In the short run, richer people are more likely to invest in these private contingencies, this means that the poor are exposed to more adaptation risk.
Case #2
Unlike in Case #1, everyone in Phoenix does not know the probability of a blackout. They know that they do not know this key parameter. The most pessimistic could actually move away from Phoenix. The most optimistic will be less likely to invest in backup power generators because they assume the probability of system failure equals zero. This subset of people is at risk because they won’t have invested in an adaptation backup plan.
If there are trusted information providers, they can educate the Phoenix public about rising heat and rising grid stability risk. Such an education would lead a Bayesian to update her subjective beliefs about grid stability and she will be more likely to invest in self protection.
A FINAL POINT
I have presented this adaptation challenge from the perspective of a given person living in Phoenix. They can move away to a cooler place, they can invest in backup power. Now let me turn to the power provider.
I have recently written a piece about how to use dynamic pricing to keep the grid up and running. Read this to see an optimistic piece about new “rules of the game” for keeping supply and demand in balance in real time.