Energy Inefficiency, with Joshua Blonz

References and Recommendations

• The Welfare Costs of Misaligned Incentives: Energy Inefficiency and the Principal-Agent Problem by Joshua Blonz
• Lyft Doesn’t Cause Congestion, All Vehicles Do by Severin Borenstein

The Full Transcript

Daniel Raimi: Hello and welcome to Resources Radio, a weekly podcast from Resources for the Future [RFF]. I'm your host, Daniel Raimi.

This week, we talk with Josh Blonz, a postdoctoral fellow at RFF. I'll talk with Josh about his recent research into an energy efficiency program in California. We'll talk about the goals of the program and its outcomes, which are not always good because of something economists like to call the principal-agent problem. Josh will describe what that means for policymaking on energy efficiency, your next trip to the auto mechanic, and much more. Stay with us.

Josh Blonz, thank you so much for joining us today on Resources Radio.

Josh Blonz: Of course. Happy to be here. Thanks for having me.

Daniel Raimi: Absolutely. So, Josh, we're going to talk about a recent paper that you've written, but before we get into the details on that paper, we always like to learn how our guests got interested in energy and environmental topics in the first place. So, can you tell us a little bit about what piqued your interest in this world of research?

Josh Blonz: Sure. So, actually, my background is kind of a very RFF-centric story. When I graduated from undergrad as an economics major, I got a job as a research assistant working at RFF, working with Dallas Burtraw and Margaret Walls on some energy and environmental questions, and I really enjoyed it. I decided to go to grad school and kind of continue studying energy and environment. I went to Berkeley, where they have quite a nice program looking at these kind of questions, and upon graduation came back to RFF because I liked it so much and really enjoyed the topic area.

Daniel Raimi: Fantastic. So, RFF through and through.

Josh Blonz: Yep.

Daniel Raimi: That's really great. So let's get into this paper. The paper we're going to talk about, so people can find it, is “The Welfare Costs of Misaligned Incentives: Energy Inefficiency and the Principal-Agent Problem.” So we're going to unpack in the next 25 or 30 minutes. The paper focuses on an energy efficiency program in California. Can you tell us a little bit about energy efficiency programs, broadly, perhaps at the federal level, and what some of the economics literature has told us about those programs over the last several years?

Josh Blonz: So, energy efficiency policy is this incredibly popular policy tool. It even has a little bit of bipartisan support because it's seen as this win-win policy scenario where maybe your local utility, or your state government, or the federal government will provide some subsidies to people to increase the efficiency with which they consume energy. So, think of weatherizing your home, or making sure you have really new energy-efficient appliances. Maybe you've seen the Energy Star program out there. This is kind of one of those energy efficiency programs. We're spending 8.2 billion dollars a year on these types of programs, and the reason is because it seems so attractive—that it will help people save on their monthly utility bills, which is kind of a little bit of a government financial push. Everyone's going to be better off because they will save money on their local utility bill, and we're going to reduce energy consumption, which is going to help with CO2 [carbon dioxide] emissions [and] local air pollution.

It sounds great, but what the recent economics literature has really found is that these programs are underperforming. They promise to save you a hundred dollars a month on your electricity bill, and they're saving you maybe 30 or 40 dollars. They're not delivering on their promises, and when they don't do that they become kind of ineffective. They become not a good use of our tax dollars, and it's been a bit of an open question as to why these programs are underperforming. There've been some papers which have kind of put forward a couple of hypotheses—things like the models that are used to predict the savings are a little bit optimistic, or maybe people are consuming more energy once it's cheaper to consume more energy. But there isn't really a concrete answer. It's potentially a mix of many different factors. One of the things I'm doing in this paper is addressing this question head-on and thinking about some of the reasons why this might be happening.

Daniel Raimi: Right. So we're going to get into that. One quick clarification question—you mentioned 8.2 billion dollars in expenditures. Is that just the federal government, or is that the sum of federal, state, and local other initiatives?

Josh Blonz: That's federal, state, and local. That's all of it. A lot of it is going to be run kind of at the local level, like the public utility commission in a state will direct the local utility to run a program. They'll raise the money, kind of on people's electricity bills. If you looked at your electricity bill, there might be a little charge on there that says for energy efficiency programs. Maybe not that straightforward, but that's what the money's going to.

Daniel Raimi: Right. Let's start talking about this paper. As I mentioned, the phrase “principal-agent problem” is in the title. So, can you tell us briefly what the principal-agent problem is, kind of in general terms, and what sorts of issues it tends to create in an energy and environmental context?

Josh Blonz: Of course. So, the principal-agent problem is definitely a bit of economics jargon, but it really captures this idea that I think most people deal with on a day-to-day basis in their everyday lives. So, a principal-agent problem is when you, as the principal, hire someone as an agent to do something for you. Maybe it's your car mechanic. You need someone to service your car, or something's not working, and so you take your car into the mechanic and you're asking this person to diagnose what's wrong with your car, and then tell you what that is, and then you're going to pay them to fix it.

Now, the idea of the principal-agent [problem] is that you, as the car owner, are going to have maybe some different incentives from the mechanic. The mechanic might want to make some money on this job, or make more money by replacing something that is larger, or kind of a bigger repair than you might need. The idea of the principal-agent problem is you never really know exactly what this person's incentives are.And, in some cases, it can lead to some kind of bad outcomes.

Daniel Raimi: As a total ignoramus when it comes to car issues, I am sympathizing along with your story, and imagining the many times in which the bad outcomes have occurred in my particular case.

Josh Blonz: Yeah, there's some interesting research papers out there that would show that mechanics like to sometimes go for the more expensive repair, when given the choice. So I take this kind of everyday principal-agent concept, and I apply it to these energy efficiency programs. The way I'd like to think about it is—who is actually providing these energy efficiency retrofits? Let's think about contractors that come into someone's home to put new insulation in or provide them with some new appliances. These contractors, themselves, have their own sets of incentives. They're responsible for implementing these programs, and so it's important to consider what are these contractors interested in and how that might affect the overall performance of one of these energy efficiency programs.

Daniel Raimi: Right. So let's look at one of these particular subsidy programs that you focus on in the paper, which is a program in California that has to do with refrigerators and other energy efficiency measures (but specifically refrigerators is what I think you focus on). So, can you tell us a little bit about this particular program, and kind of how much it costs for the state, in terms of either government outlays or tax expenditures, and what its goals are?

Josh Blonz: Sure. It's called the California Energy Savings Assistance Program. It's a bit of a mouthful, but its goal is to help people in California who are low-income deal with the high energy bills that we have in California. It's a pretty expensive state to purchase electricity in, and the idea here is it's a free program—so eligible low-income households can sign up and the utility will come into their home, determine what sort of upgrades they might be eligible for, and then provide them at no cost to the homeowner. So, it can be a really great program if you're low-income. Around a third of the state qualifies. They spend around 285 million dollars a year on it in California, so it's actually a pretty substantial program that's supposed to help low-income people.

So I focus on, like you mentioned, this refrigerator replacement component of the program. One of the reasons I was drawn to this part of it is it's kind of one of the most important parts of the program, based on the people involved with running it. They say that this is where they're getting most of their energy savings from in this policy. The way they've structured it is if your refrigerator, when you sign up for the program, was manufactured in the year 1992 or earlier, you as the homeowner or the renter (the person who lives there) is going to get a brand new, free, energy efficient refrigerator. They're going to take out your old one. They're going to dispose of it. They're going to give you this brand new refrigerator.

Daniel Raimi: That sounds good.

Josh Blonz: Yeah. If you're low-income in California, this is a really great program. It's designed to help people who are renters as well. So a pretty common problem is—you're renting, you're paying for your energy bill, but your landlord may not want to give you the newest appliances because they're not paying the high energy bills responsible for the old appliances. So this program can kind of come in and just give you that brand new appliance and help you save on your utility bill.

Daniel Raimi: So, people in California who are taking advantage of this program—they are presumably saving some money on their energy bills. But you identify a principal-agent problem that manifests itself in the form of these refrigerator replacements, particularly looking at elements of the contract structure for the contractors who come in and do the refrigerator replacements. Can you talk a little bit more about how the principal-agent problem manifests itself in this context, and how the contract side of things plays into this story?

Josh Blonz: Of course. Contractors in this program are being paid for the amount of work they do. So, they get paid for replacing the refrigerator, enrolling people in the program, and all of it. This is a pretty typical contact structure. If you've ever had someone work on your house, you pay them for the job they do. But it also gives these contractors this incentive to maybe break the program rules a little bit and actually misreport data. If they walk into someone's home, they can look at the refrigerator. There's usually a little plaque on the inside of your refrigerator that tells you the serial number, from which you can very easily determine the age it was manufactured. A contractor can come into someone's home, they can see that a refrigerator was manufactured in 1995, which would make it ineligible under this program. But they can just write down 1992 in the program documents—and that refrigerator is going to be replaced under the program.

These contractors have the incentive to do this because they're going to be paid to actually do that replacement. So there's a very large incentive for them to do this misreporting, and this is paired with the fact that the utility isn't verifying the data that they’re submitting—so it actually kind of makes it an easy situation for these contractors to just tweak the data a little bit to increase their own compensation.

Daniel Raimi: Yeah, it would be hard for the utility to track down every refrigerator at the dump and track it to each document, right? With this program?

Josh Blonz: Yeah. There are monitoring systems that you can imagine implementing. You can have the contractor take a picture of the little plaque on the inside with their cell phone and submit that as a part of the packet. For me, as the researcher, it's actually quite nice that they don't have that part of the program there because I can see the incentives of these contractors and how that affects their reporting behavior.

Daniel Raimi: So, as policy researchers, we want to try to fix policies, but not too much. We have to keep a little bit of errors out there so we can keep finding things to work on.

Tell us a little bit about what you found when you looked at the principal-agent problem in this program. How does it affect energy use and other outcomes, like overall welfare, from the perspective of the homeowner on energy use or from the perspective of society when it comes to overall welfare?And what are the costs of this type of misreporting behavior by contractors?

Josh Blonz: First, what I end up finding is that around half of the refrigerators that were replaced in this program were done as a result of this misreporting. So it's actually a pretty large deal. It has some pretty significant effects on the overall program. To quantify, actually, how this matters, I use household-level electricity consumption data so I can observe—before and after—these households' consumption. Using some econometric techniques, I can estimate what are the effects of a refrigerator replacement, and I can do this differently for those that qualified for the program, the ones that they were intending to be done, and the ones that didn't qualify. What I find is that the qualified replacements (those manufactured before that cutoff year), they're saving households 16 percent of their monthly electricity consumption, which is quite a lot. Old refrigerators used a lot of electricity, and so having a brand new refrigerator is actually going to do a pretty good job at saving electricity.

In contrast, the unqualified replacements save half as much. They save only 8 percent of a household’s monthly electricity consumption.So there's this very large difference between the ones that followed the program rules and the ones that didn't. The forces that are driving that is that they designed this program very intentionally to remove the oldest, least-efficient refrigerators. And starting in 1993, refrigerators became quite a bit more efficient because they have to satisfy some federal guidelines [and] some minimum efficiency standards.So what these contractors are doing by breaking program rules is that they're replacing these much newer and more efficient refrigerators, and hence getting a lot less energy savings out of the program for the exact same cost.

Daniel Raimi: Right.

Josh Blonz: So, then, I can take these electricity savings numbers and think about what us economists like to call “the welfare effects,” which is kind of a jargon-y way to say “the social scorecard.” I'm the social planner who wants to make everyone better off. I'm weighing all these different factors to go into this welfare effect, and I make this calculation using all these parameters I've thought about in this paper.What I find is that qualified replacements—those that follow the program rules—they're increasing welfare by $60 per replacement. This is going to include a whole lot of things. This includes externality reductions, so we have lower amounts CO2 emissions [and] local air pollution.Those are all great benefits associated with making these replacements. We're making the households better off. They're paying less money for their utility bills. That's another really good benefit to have. There are costs associated—you have to purchase these refrigerators. You have to pay people to install them. So that all goes into the calculation.

Now, when I do that same calculation for these unqualified refrigerator replacements, I find that they actually reduce welfare—on the order of $106 per installation. So there's a really big difference between those that followed the program rules and those that didn't. The way I like to think about all these results together is from the policy perspective. You can take a program, which on paper appears to be welfare improving. It's making everybody better off, when you take into account the cost and the benefits, but when you introduce the incentives of these agents to the situation, these contractors, you can actually turn the program into a welfare reducing program. This might be one of the explanations for why these energy efficiency programs are underperforming. That the people tasked with doing these installations, with these retrofits, are actually making it less effective.

Daniel Raimi: In the paper, and you mentioned this early on when we were talking a few minutes ago that there are these numerous papers in the economics literature finding that these energy efficiency programs are less effective than we might hope that they are. So, it sounds like your paper is explaining, at least potentially explaining, part of the reason for that ineffectiveness. So do you see this as a potential avenue to maybe improve these efficiency programs going forward, or are there other solutions out there that you're thinking about?

Josh Blonz: Exactly. So most energy efficiency programs use this third party contractor structure, and so it's likely that other energy efficiency programs are susceptible to these exact same kind of principal-agent problems that could be reducing the effectiveness of the programs. Now it's going to be context specific but, I think, what one of the findings in my paper suggests, is that we really need to be cognizant of these incentives. It hasn't really been talked about and addressed directly, and so I think policymakers, as they continue to implement these energy efficiency programs, should put monitoring into place—or think about how the contract structures might be incentivizing agents to maximize their own payouts, which could potentially hurt the overall program outcomes.

Daniel Raimi: And that could apply at all sorts of levels, right? Federal level, state level, as well as private utilities that are undertaking these efforts on their own. Are there implications of this work for other non-energy industries?

Josh Blonz: There are many kind of papers that have looked at these types of principal-agent problems. Bonuses are a big area. So, people trying to reach their year-end bonus by hitting specific sales targets.And there have been some papers out there that have shown how people are strategically reporting their bonuses in certain ways, using accounting tricks to hit their bonus. It's kind of unclear as to what are the actual consequences of these types of behaviors. So what my paper is doing is picking one of these principal-agent cases, and showing that there are some really significant welfare costs associated with it.

Now the principal-agent problem isn't this kind of new idea. It's pretty well understood in the economics profession, but understanding what the consequences are is really important, and I think there's a lot of research to be done using this framework to examine various federal programs, state programs, or just kind of contracts in general between employers and employees.

Daniel Raimi: Right. Yeah. Fascinating. At RFF, we don't have bonuses, so I'm not asking my questions in any different way than I otherwise would. Or at least I don't get any bonuses. If other people do, I'd like to know about it.

This is really fascinating work but I know you have other interests when it comes to the world of energy and environments—not just refrigerators, and energy efficiency, and principal-agent problems. What are you interested in and working on now?

Josh Blonz: I'm working on this new project with a number of coauthors at Resources for the Future, where we're investigating how these new “smart” home energy saving technologies can actually benefit consumers. Now there's been some studies out there looking at how smart thermostats can make people better off. I think these are really important questions—but how most of the literature addresses these questions is by looking at how these technologies can affect energy consumption. Do people reduce their energy consumption when they get a new thermostat? That's one way to look at it. But it can miss one of the important parts of the equation, which is, how comfortable are people in their own homes?

One way that you can conceivably save energy is just turn off your air conditioner in the summer. You'll sweat a lot but your utility bill will go down quite a bit. It's kind of unclear whether that has made you actually better off as a consumer. We try to get at this question by using data from one of these new smart thermostats, and it kind of creates this new stream of data that allows us to observe things we've never been able to see before. So we have data on when people are home, with motion sensor data. We know what their preferred temperature is inside, and we know what the actual temperature is at these five minute intervals. We can use this data to evaluate one of these smart scheduling features that comes on these thermostats to see if it actually makes people better off. What we find is it actually does help—to some extent.

Our preliminary results show that these technologies can allow people to save energy while maintaining the same level of comfort. So they're just as comfortable, but they're using less energy. This is great. This is exactly what we're after as economists, and as policymakers.

The [results] suggest that these types of data, these new technologies, both can help people out and also allow us to learn something about how consumers are interacting with their appliances—how they're consuming energy at this disaggregated level. Before, we had meter-level data. We might even know at 15-minute intervals how people are consuming energy, but we don't know how much of that's going to the refrigerator, or to the thermostat, or to the lights. So this data allows us to get in there and really understand what's going on. I'm really kind of excited to work in this area.

Daniel Raimi: Yeah, that's great—all these new data streams that are available from the proliferation of technology around the home. I know there are other researchers interested in this stuff, and I'll be really interested to read the results of your work on this thermostat issue when it's ready.

We're going to close it up now but before we go, I want to ask you the same thing that we ask all of our guests, which is: what's at the top of your literal or metaphorical reading stack—or maybe at the top of your tabs on your internet browser? What's something that you've read, or watched, or heard recently related to the energy environment that you find really interesting that you think our listeners would be interested in as well?

Josh Blonz: I read a really great blog post this week that the Energy Institute at UC Berkeley puts out, which I would recommend to all the listeners to check out, written by Severin Borenstein.And it's looking at the idea of ride sharing, and how that affects traffic. I'm sure everyone has been using Uber and Lyft. It makes our lives quite a bit better, but people also really don't like sitting in traffic. So there've been some studies that have shown that maybe these types of ride-sharing platforms are contributing to traffic. This is upsetting a lot of people, especially the people who own cars and sit in traffic all the time. Some cities have started to take some actions to deal with this problem. New York is a prominent example. They're trying to cap the number of Uber drivers that are out there. Which, as an economist, this isn't really the right way to deal with this kind of situation—because, if you think about it, the problem itself is not the ride sharing. The problem is the traffic itself.

Economists have long suggested that the best way to deal with this is by just directly taxing the externality itself, which is congestion. There are a lot of ways to do this. You can charge people who drive into traffic-y urban areas during rush hour, or charge people to drive on freeways that are usually very congested, and this way it will kind of give people the right price for their driving. These bans, these caps on the amount of ride-sharing drivers maybe are not the right way to go about it because that might help out the people who previously owned cars and were driving—but it really kind of shuts down the opportunities for new transit.

These types of ride-sharing apps, as we move into the future, are going to be changing dramatically. Self-driving cars are what everyone loves to talk about. They're really going to require having the right policy prescriptions to deal with the congestion they might charge, and I think this blog post does a really nice job of talking through either side of it—and then really suggesting that the economic solution would be the right way to go about it, which is taxing congestion. There are all these added benefits. You can use all this revenue you raised to subsidize public transit to help other people have alternatives to maybe not travel on roads during the most congested times.

Daniel Raimi: Right. So, sometimes congestion pricing, I think, is a broad term that gets applied to these things. What's the name of the blog post, or the name of the Berkeley blog itself, so people can look it up?

Josh Blonz: It's called the Energy Institute Blog. The blog post is, “Lyft doesn't cause congestion, all vehicles do.”

Daniel Raimi: All right—well, we'll have to check that out, Josh. Thank you so much for the recommendation, and thank you so much for taking the time to share the results of this really fascinating research on the principal-agent problem in the context of energy efficiency programs. We really appreciate you being here.

Josh Blonz: Thank you very much for having me. It was a lot of fun.

Daniel Raimi: Thank you so much for joining us on Resources Radio. We'd love to hear what you think, so please rate us on iTunes, or leave us a review—it helps us spread the word. Also, feel free to send us your suggestions for future episodes. Resources Radio is a podcast from Resources for the Future. RFF is an independent, nonprofit research institution in Washington, DC. Our mission is to improve environmental, energy, and natural resource decisions through impartial economic research and policy engagement. Learn more about us at rff.org.

The views expressed on this podcast are solely those of the participants. They do not necessarily represent the views of Resources for the Future, which does not take institutional positions on public policies. Resources Radio is produced by Kate Petersen, with music by Daniel Raimi. Join us next week for another episode.