Kip Viscusi on the Value of a Statistical Life and Coronavirus

Tom Lenard: Hello and welcome back to TPI’s podcast Two Think Minimum. It’s Friday, April 10th, 2020 and I’m Tom Lenard, president emeritus and senior fellow at the Technology Policy Institute. I’m joined by Scott Wallsten, TPI’s president. Today we are delighted to talk to Kip Viscusi. Kip is University distinguished professor at Vanderbilt with appointments in the economics department, the management school and the law school. He previously was Kogan professor of law and economics and director of the program on empirical studies at Harvard Law School and has held professorships at Northwestern and Duke. He is the author of more than 30 books and 370 articles and his most recent book is Pricing Lives: Guideposts for a Safer Society from Princeton University Press. I first got to know Kip when he was deputy director of the Council on Wage and Price Stability way back during the Carter administration. Seems a long time ago. I was an economist there. CWPS as it was known, was responsible for White House oversight of new federal regulations and was the precursor to the current Office of Information and Regulatory Affairs at OMB. Kip is a leading authority on benefit-cost analysis and is widely recognized for having done pioneering work on valuing risks to life and health and his estimates are currently used throughout the federal government. It’s a pleasure to have you here, Kip. Why don’t you start out by telling us how your estimates are used by the federal government?

Kip Viscusi: Well, mostly they’re used in regulatory impact analyses. Government agencies are required to assess the benefits and costs of regulation and to make them comparable, often you have to convert the benefits into dollar terms. The main benefit from health, safety, and environmental regulations typically consists of the mortality risk reduction associated with these regulations. So to monetize these reduced mortality risks, government agencies have been using these estimates of value of a statistical life, to translate reductions in expected number of deaths into some kind of dollar equivalent.

Tom: So explain the concept of the value of a statistical life.

Kip: Well, to monetize the expected lives that are saved through a regulation, government agencies used to look at the earnings people had or the medical costs associated with the death, but that was divorced than the basic principle of how we should think about benefits, which is society’s willingness to pay for the benefit. So in this case, the willingness to pay is how much are you willing to pay for reduction in the probability of death as a result of the regulation. So to get a hand on that, what I’ve done is look at labor market decisions by workers. How much money do they require to face extra risk on the job? For example, if a worker faces an annual fatality risk of one chance in 10,000 and if there are 10,000 workers like that, then on average one of them would die. And if these 10,000 workers each gets paid $1,000 to face that extra risk, then the total compensation you have to pay these 10,000 workers is 10,000 times $1,000 each or a total of $10 million to incur this expected single fatality to their group. That’s what the value of statistical life is. How much money do people require per expected death in terms of compensation?

Tom: So when this is used in cost benefit analysis, is it controversial, that people think, is it moral to trade dollars for lives?

Kip: Well, it used to be more controversial. In fact, it was so controversial that back when we were working together, I suggested to OSHA that they adopt this approach. And what they suggested was, no, we can’t possibly do that. That would be immoral. So instead of using these values of statistical life numbers, they used what they called the cost of death, so far be it from us to value lives. We’ll just say the loss is the cost of death which in that case was the value of lost earnings. So in effect, they used a really small number. So yes, it was controversial. It probably is controversial among non-economists. So government agencies have been doing this for almost 40 years now and economists are fine with it, but typically about once every year or once every two years, the press discovers, oh my god, they’re valuing lives and does a big stir in terms of the controversy. And then it settles down when they hear how big the number is. So when they find out that their lives are being valued at a quite high number, people get less upset. Now there maybe a current controversy, some people may complain my numbers are too high. So people that don’t like government regulations want the number to be even smaller than what I suggest.

Scott: Well this seems to be one of those times when with the coronavirus that we have to think about that number again and how to make choices about how we lock down society, what we do with the economy. And people seem to be resistant to making that explicit calculation.

Kip: I have no hesitancy in doing so because we do this all the time. I don’t think coronavirus is any different than the other policies. It’s certainly a large scale. So in terms of the social distancing, you may save a million lives. Well, that’s certainly a whole lot more lives than are typically saved through government regulation. But that doesn’t mean you shouldn’t think about it the same way. And just like we try and figure out what’s the most effective way to design a government regulation, you should also try and figure out what’s the most effective way to design a response to the coronavirus crisis.

Scott: Do you think policy makers should be taking that more into account or do you think that they are and it just hasn’t entered really into the public debate so much?

Kip Well, right now, at least in terms of taking initial action, I’d say it’s a slam dunk that doing something is desirable.

Scott: Absolutely.

Kip: So you know, you talk about a million lines. If we’re really saving a million lives through social distancing, multiply that by $10 million a life. You’re talking about $10 trillion in terms of benefits. That’s a lot of benefits that we haven’t even gotten to the illnesses that are unpleasant and costly. So the idea that you should do something now, if you were to think like an economist, the answer would be yes, but sooner or later we’re going to have to be phasing down and the government’s going to have to start picking their shots. You’re not going to want it continue social distancing throughout the rest of our lifetime. And what are the things you bring back? What policy measures do you adopt? Thinking about that, and it involves the same kinds of thinking about benefits and costs and marginal benefits and marginal costs that you do for any government policy.

Tom: So using the example you were using, which probably is a somewhat realistic example about saving, let’s say saving a million lives, compared to if we did nothing, and if you used your $10 million per statistical life saved, that means we should be willing to spend $10 trillion, which is not too far from cutting GDP in half. And hopefully we will do better than that. We’ll take our big hit to GDP, but it’s not going to be half, hopefully. So we’re at least passing a cost benefit test if we’re not passing a cost effectiveness test.

Kip: I agree. I think this certainly passes the test that the benefits exceeded the costs. We’re doing something and whether they’ve picked the best policy is unclear. Since law, all different state laws, states have different policies and different levels of restrictions. Where I am now in Colorado is more restrictive than where my son is in Georgia with the beaches are open and is much less stringent in terms of the lockdown procedures.

Tom: This is a little bit of a drilling down, a little bit deeper maybe than most people want to, but you know, old people are more susceptible from the coronavirus. Do we value their lives the same a young people’s lives?

Kip: Well, I don’t think this is a too deep a dive actually since I think thinking about the age distribution is something that you’d want to do for any government policy just to make sure you’re thinking about it the right way. Government agencies generally don’t make any distinctions based on age, except for people who have very short remaining periods of life. And for that, instead of using the value of statistic life, so saying that people’s lives are worth $10 million or in the case of most agencies, between nine and $10 million, you’d use the value of a statistical life year. So if you’re only losing a year of life or in some cases a month of life, then you’d want to scale down the numbers and the current numbers for the value of statistical life, you’re on the order of $500,000, which is still a fairly large number, but 500,000 is a whole lot less than 10 million. So yes, the age distribution could matter, as might if people are seriously ill and they’re not in their nineties or they have long life expectancy if they had not been sick, but if they’re seriously ill and only have the short period of remaining life. So it’s not just age, it’s the future of life expectancy you’d want to look at and take that into account.

Tom: In terms of different countries, this is supposed to hit poorer countries very hard. Maybe not quite yet, but over time, less developed countries. I assume that the value of a statistical life is correlated with income.

Kip: Well we don’t even have to assume it, it is. So we’ve done estimates across countries and there’s roughly a inelasticity of 1.0 so a 10% increase in your income increases your value of statistical life by 10% or, similarly, since everybody else, other than maybe Qatar, is poorer than the United States, a 10% decrease in your income implies a lower value of statistical life. So throughout the world, if you’re looking at the UK, if you’re looking at the many hotspots, what we’ve had for coronavirus, if you’re looking at Spain, Italy, China, their value of statistical life would be less than that of the United States. And that’s because based on their income, they have different preferences for risk versus money, how much they’re willing to spend on greater amounts of safety. So this is a reflection of their preferences. It’s not we’re saying their lives are worth less from our standpoint, we’re saying this is how they think about safety and their policies should reflect their own preferences, which in fact they do.

Tom: So what about also the steps we’ve taken? You know, shut down significant amount of economic activity. I mean that also has health and mortality facts. How should we be taking those into account?

Kip: Well there’s actually a whole literature that I’ve contributed to on that as well in terms of the relationship between income and mortality. So the richer you are, the healthier you are, generally. And during times of economic downturn, increased unemployment, mortality rates go up. What I’ve found is for a loss of less than $100 million, anytime we lose that amount of money, there’s one expected death, across the economy. So if Bill Gates lost $100 million, I’m sure he’d be fine. But in terms of the economy, more generally, if you lose $100 million, that kind of loss distributed across the economy, it tends to lead to one death. And the reasons for that includes suicide by people who’ve lost their job. And when people have less money, they have less money for safe food, safe housing, and so on, all the way down the line in terms of their consumption behavior. So when we think about the economic costs associated with these policies, we should also recognize that we’re not just talking about money. If you’re imposing a huge economic cost, there’s also a health loss associated with that. So health is on both sides of the analysis.

Scott: In this case, there’s also this, well maybe it wasn’t unexpected for everybody, but there are fewer car accidents and fewer other health incidents, because people just aren’t doing things. Of course, we’re getting those benefits of the short run and the costs that you’re talking about take place over the longer run, right? So we have to discount those.

Kip: That’s right. So the costs are not necessarily immediate. There may be long-term effects of bad nutrition or living in an unsafe neighborhood because you don’t have the money to live in a safer neighborhood. And we do get some benefit. You mentioned decreased car accidents. I’m getting a 15% kickback on my auto insurance rates because of this. 

Scott: I think I need to change providers.

Kip: At least from one of my insurers. I’m not sure if the other one’s given me that kind of kickback, but that’s at least providing some modest financial benefit during this period.

Tom: And also, maybe should have asked you this before, but there’s all sorts of costs going on. The federal government spending a lot of money. We’ll talk about the loss of GDP because of closing down a portion of the economy. What is the appropriate measure of cost?

Kip: It’s very interesting. I think you want to look at the kinds of counting of costs that the Office of Management and Budget would generally do with respect to regulations. So we care about the net dead weight costs to the economy. So if we’re looking at costs that are like transfers, passing money around, those don’t count as net cost or net losses. But if you’re talking about people losing their jobs and restaurants not selling anything or hiring anybody to do the work, those are really economic costs. So you want to tried to distinguish the transfer payments which are passing money around as opposed to the actual economic losses that are taking place. The $2 trillion, for example, a huge part of that’s transfers. It’s not a dead weight economic loss to society.

Tom: So would an approximation of the real economic costs, it’s not quite precise, but the best thing we have be the loss of GDP?

Kip: If I were to start with a number, that would be a good one. For one thing, you can get a handle on it. Building things up piecemeal is probably harder and this is the number you’ve got. So yeah, I think that’d be a great number to start with.

Scott: You’ve been more successful, I think, than, arguably, almost more than any other economists in having your work become part of how policy is decided and these ideas brought into practical use by the government, not just ours but around the world. Are there areas that you’re disappointed with that you think haven’t used these analyses, this approach, the way you think it should be done or places that haven’t adopted it and therefore are making poor decisions than they might have otherwise?

Kip: Well, right now I’d say this approach or the adoption of this approach is the dominant approach for assessing new government regulations. Whole lot of other uses that you could make in terms of thinking about the value of mortality risks. So when you think about where are the sanctions for violating government regulations? So if you have a company that violates safety standards or environmental standards, as a result of that, there’s a large number of deaths. Well, how do you figure out the sanctions for them? Well, currently the sanctions are governed by statutes that really limited to trivial amounts. I mean, the level of the penalties is really small. So there’s a real imbalance between the kinds of sanctions the government levies for the violations as opposed to how they think about the lives and their value when they’re protecting them through proposed regulations. Also, I think companies should use these numbers as well. And understandably, companies are a bit gun shy on this front. Years ago, particularly in the auto industry, companies would do the exact kind of thing economists would want them to do. They think about the benefits of safety regulations or new safety devices for cars, and they think about the costs. However, when they tried to monetize the benefits, they looked at the court awards in the case of wrongful death cases. And those numbers tended to be fairly low. And when it became public that yes, GM or Ford or Chrysler, all three of them, would value lives lost based on how much they’d have to pay off in court if the person was killed, then people got upset and they got really clobbered with big punitive damages awards. So one way to give an incentive for companies to do this, would it be to also give them legal protections so that it couldn’t be used against them if they did or if they did the equivalent of a regulatory analysis for safety devices for cars using these value of statistical life numbers, and you can’t introduce that as evidence to try and show that they were reckless and irresponsible and should be subject to punitive damages, which is what’s going on now. So yes, companies should do it, but they need to have some inducement or legal protection to enable them to do it without getting clobbered.

Scott: I mean that sounds like it takes away the reason to avoid doing it. How does it give them an incentive to do it, to take the societal effect, the values of statistical life into account, that doesn’t seemed to help them.

Kip: If they’re trying to figure out how much do consumers value safety, which is where these numbers come from, so it’s not just workers. These same kinds of estimates you can develop based on how much consumers are willing to pay with for cars that have airbags when they first came out and you get numbers in the same ballpark. There’s a way for them to integrate into their thinking about safety, what the value consumers place on the safety devices are and doing it in an actual meaningful way as opposed to guessing, well, this might be a good idea. It might not, but if they could actually ground it and say if consumers understood what this did for them, this is something they would want, then that would be I think a relatively sound approach to making corporate design decisions.

Scott: So it would take into account what people care about for safety rather than whatever tools they use today.

Kip: Right. And both in respect to the product design decisions as well as government policy as well as health policy, I think economists should play a big role, not turn everything over to engineers and turn it over to medical experts. A lot of the battles in terms of setting government regulations are between economists and scientists, so they may have different agendas or different ways of thinking about it. So we did studies, hazardous waste clean-up programs, the superfund programs for EPA. And there was a constant battle between us and the superfund office people because they just didn’t think like economists in terms of how they should do it. So they had sites where they were spending a trillion dollars per cancer case prevented. That’s not because they spent that much money. That’s because the denominator, it was so small that the amount of money’s spent per cancer case was like astronomical. So they just weren’t thinking like economists and they were allocating resources totally inefficiently. We found that for their current budget, they could have cleaned up over 95% of all the hazards waste sites across the United States had they spent the money effectively and targeted where the people are, which was not anything that they were used to thinking about. So having economists involved and making people think like economists would make not only regulations more effective, but also make products safer and save money, probably, as well. 

Tom: If you were asked by the folks in various states or the federal government, how quickly they should start opening up economic activity, what would you tell them? What type of analysis would you tell them to do or would you do to form their decisions?

Kip: Well, first I’ve been disappointed that not all States have been equally vigorous and that wouldn’t be a problem except people travel. So, for example, once I return back to Tennessee, Tennessee is one of the not as vigilant a state as some of the other ones, and Kentucky, which has very strong lockdown restrictions. They’re sort of leery of people driving over from the state of Tennessee because they haven’t been subject to the same kind of lockdowns. So ideally it’d be great if we could achieve a low level across the United States, low incidence rate across to the United States, so that if people did travel we’re not at the risk of basically restarting this again, and resort areas are particularly vulnerable. So I’m not in Vail, I’m in a different part of Colorado, but the Vail area has had hundreds, nobody lives there. It’s a very sparsely populated area. They have hundreds of coronavirus cases because the incoming skiers from New York and other areas have brought it with them and it’s taken off. So unless you have a very effective national crackdown, I think it’s going to be pretty hard to get rid of. But then you can start thinking about which kinds of activities to bring back. And originally out here, they banned, like bike repair shops were excluded. Now they change that and now they’re back in. So I think they’ll probably go industry by industry and try to figure out where can we actually have activity brought back and not have a lot of social contact that’s going to increase the risk. Now there’s some areas where you’d not necessarily have to have contact, but people value it. So men and women are going to go without haircuts and without going to the hairdresser for women, I don’t think that’s going to go on for the rest of the year. I mean there’s no question about that.

Scott: Tom already looks like a hippie.

Kip: Right. So I think if you don’t start relaxing some of these things, people may just say the heck with it. I’m not going to play this game any longer. And you want to maintain enough discipline so that if people look at their daily lives and say, this is something I really value. Yeah, I know my hairdresser is going to take extra precautions. They’re going to sterilize the door knob, keep the door open, maybe so nobody has to get contaminated going in, clean up the facility after every day, et cetera. Then I think that certain things will start to be brought back but with extra precautions and once again you’d focus on the ones for which people perceive the greatest benefits relative to the costs.

Tom: Do you think the numbers, I mean, in contrast to many or at least to some regulations, there doesn’t seem to be, in this case, we don’t seem to be doing values of the benefits, the values of statistical life, we don’t seem to be doing too much. I get the feeling that you think that if anything we’re doing too little. Is that, am I interpreting that correctly?

Kip: I’d say some areas are doing too little. So if you say it’s okay to go to the beaches, I think that’s too little. And I think a lot of it’s inconsistent and it’s not clear that the inconsistency is because of differences in risk levels in the different areas as opposed to different political attitudes. I guess the attorney general referred to these measures as draconian. Well, sure. They may be a smart thing to do at the current time, to have a harsh crackdown. In general, I’m not that happy with the national policies, but I think a lot of the States are doing a great job.

Tom: And in terms of opening up, obviously it may depend state-by-state, it depends. With opening up, if you had a bias it would be to do with it slower rather than faster, am I correct in suggesting that?

Kip: I’m not sure it’s so much slower or faster. I wouldn’t turn everybody loose yet, so I’m not rebooking my Disney world trip, for example, and I’m not sure that the Rolling Stones tour is going to happen where they have 70,000 people in the stadium at the same time and, and the opening day for the baseball season may not be in this calendar year. I think certain things will come back and I think you should probably do it in terms of thinking sooner rather than later. I think in terms of relative cost effectiveness or relative benefits and costs in the different areas that we can bring back and target or bringing back in the areas where you get the greatest bang for the buck and that may not necessarily be financial, it could be these things that people really value and want and are important to their lives. And if they’re doing that they’ll take extra precautions. Like we do things now, I go to the store once a week and so it’s not like I’m locked up in my room, hermetically sealed and not exposed to any risks. So there are certain things we do do now that are not risk-free and we decide that the benefits of going to the store exceed the costs and it’s the same kind of thing. As the risk diminishes, there’s going to be more kinds of things that kick in. And I think people are really learning that precautions matter and hopefully they’ll continue to take precautions once we reopen these different activities.

Tom: There seems to be, certainly a view in the press, and maybe much more generally, we were not, we at least in the United States, we did not prepare adequately for this. How do you think societies should in general plan for low probability, high costs events of uncertain timing?

Kip: Well, the high cost is very high and even if they understood the probabilities, I don’t think people prepare enough. So I was on a dissertation committee and I’m coauthor of a paper coming out of it after a previous virus incident. And should we invest money to save lives based on a pandemic or should we invest money to save lives from terrorist attacks? If you ask people on the street, most people don’t care about pandemics, it’s just not real. Whereas terrorist attacks are flamboyant and real, and they could anticipate that. So I think there’s very little anticipation of anything of this scale.

Tom: Do you think. now it probably become more real to people?

Kip: I think now it’s much more concrete. I’ve never lived through a pandemic, a real pandemic in terms of one that affected me.

Tom: Yeah. I haven’t either. Up until now. My parents were kids during the Spanish flu, but I was thinking about the fact that they never really talked about it. So I don’t know what effect it had on them.

Kip: The great depression I had heard more about than the Spanish flu cause my parents I guess were like five and six during the Spanish flu. 

Tom: Scott, do you have anything else? I think we’re getting close to the end.

Scott: Yeah, no, this was fascinating.

Tom: Well, thank you very much for sharing your thoughts on this, Kip. Really appreciate it. We’d like to get your back sometime and discuss this after there’s some additional developments if you’re willing to do so. Thanks very much. 

Kip: Sure. I noticed nobody doing this talk today was wearing a face mask. 

Scott: Unless it was very realistic.

Kip: I think the marginal benefits of being inside wearing a face are pretty small, so I think that’s a wise style choice. 

Tom: Okay. Really hard to get too. Thanks Kip, we’ll talk to you soon.