Winston on Market Failure and Government Failure

Welcome to Econ Talk, part of the Library of Economics and Liberty. I'm your host Russ Roberts of George Mason University and Stanford University's Hoover Institution. Our website is econtalk.org, where you can subscribe, find other episodes. Comment on this podcast and find links and other information related to today's conversation. Our email address is mail at econtalk.org.

Today is december eighteenth, two thousand nine, and my guest is Clifford Winston, Senior Fellow at the Brookings Institution and the author and uh editor of many books, including what we're gonna talk about today. Government failure versus market failure. Cliff, welcome to Econ Talk. Great to be here. So your title is uh Government Failure versus Market Failure. Most people hear the term market failure. It gets thrown around a lot today, uh often to mean things like

Markets didn't do what I wanted them to do. Uh they failed me, or they failed what I'd hoped they do. And that's not what economists usually mean by market failure. We have a fairly uh narrow definition. So why don't we start with with talking about what

the standard textbook definition of market failure is and then we'll talk about government failure. Okay, sure. Well market failure, you're right, it is something that everyone has a their own uh idea about. Things didn't work out for them. But economists uh have a very precise notion of what market failure is and it and it does focus on what we call efficiency issues, that is the allocation of resources.

And effectively what we're looking at are situations where we get an equilibrium where it would be possible to make one person better off without making anybody worse off. And there's a technical uh term for that pareto optimality. But that's sort of a standard what we call market failure. That that you can sort of reallocate resources

That seems pretty uh attractive and it's named for the great I always liked his name, so I'm gonna give his full name. The Italian economist, uh late nineteenth century, I think. Uh Vilfredo Pareto. made up name, but that's his actual name. It's a great name. Love it.

So Pareto, that was one measure of how would we know that the world's better today than than another world uh that a world A is better than world B is that we'd like world A if somebody could be made better off without making s anyone else worse off.

So go ahead. Yeah, th that would seem sort of well, how how how could we ever forego any any situations like that? And And w really what we're talking about are what we would say potential Pareto improvements because usually what's necessary is to actually compensate some people um so that they're no worse off and then you still have something left over and some people are better off.

So th the notion of actual improvement is very, very rare. You can't imagine any policymaker who who would allow anything like that.

uh to remain. You'd figure, well, what's the harm of of making, you know, somebody better off and nobody else worse off. Usually there involves a transfer of resources involved and you're talking about a potential improvement where you could at least compensate people uh who who appear to lose and make them at least as well off as they were before, you'd still have something left.

And that's really what we're talking about most of the time in practice. And this is this is a a very essential concept in modern uh public finance and public wealth and welfare theory. I have to say over the years I found it increasingly um

uh I've been increasingly uneasy with it. So but and at the end of this conversation I hope we can get into some of those issues. But just to clarify what what we've said so far Pray to efficiency or Policy changes that make the economy more efficient are simply saying the pie got bigger and we're not going to worry about whether some of the people Uh because this policy change might actually be worse off as long as the gains to those who are better off

are sufficiently large that they could have been compensated, which is simply way of saying the pie is bigger, even though some people shares. And that's what we're talking about. It's just it's just trying to increase the the size of the pie. Obviously that it has implications for distribution and yeah let me just quickly preview I'm sure we'll get to this later that these things are often presented as conflicts.

But they need not be. Um, you know, i if you're inside the frontier as we put it, uh, you you often can actually get ways of expanding the pie and even cutting it up then.

In a in a way that may be more favorable to distribution. But that's that's further down the road. So so the simple point though is that Is that you would want to know something about whether we're getting the maximum uh productivity out of our economy and if if there are either laws in place that are hampering that or laws that could be put in place to improve it, we might want to go that direction.

Right. Or more to the point, I mean, one of the major responsibilities of government, right, is to produce such such improvements. In other words, where government justification in theory uh occurs in in the market is where there is this market failure. And the government is then supposed to come in and institute policies that, you know, lead us to things that

make some people better off, with in principle not making anybody worse off, and in the process expanding the pie. So that's sort of the efficiency objective of of government policy. And that's really what I wanted to assess. That's the textbook claim. Markets don't always work perfectly. And if government could improve that, that should. Correct. That's that's the theoretical justification for intervention on on the efficiency ground. So what's the government failure problem?

Okay, well you know there there are several of them uh that that people certainly are familiar with. You know, monopoly is one case. Um there you have, you know, the the abuse of market power leading to to prices above marginal cost. That creates what we call a a deadweight loss. And, you know, if if the behavior that underlies this is anti competitive This is a market failure which, you know, supposedly antitrust policy

uh is intended to correct. And so, you know, they they would presumably step in and and do something that that leads to a more competitive outcome and prices to come down uh and efficiency to be improved. You want to define a dead weight loss. Tell tell those sort of things. uh to society in the process of of maximizing their profits.

um the monopoly would reduce output and even though there are cases where on the margin you know people would like to pay the cost of the output that is produced, they're not able to and that output is just basically lost. uh to to to society. There's a potential exchange that could make the monopolist and the in theory, could make the monopolist and the buyer better off, but I mean obviously have to lower the price than all the other units.

Unless there's a way to to avoid that and that would discourage a monopolist from doing that. So there's a a foregone bet net benefit. That's the claim. So go ahead. So there's antitrust. What else? Aaron Ross Powell And then a technological one called natural monopoly, that's sort of the unusual situation Where uh output or the costs are minimized with one producer, you have declining average cost is is certainly one condition that could do it.

And the difficulty that arises there is in that in such a situation you have average cost below marginal cost. Um uh sorry, the other way around, you have marginal cost below average cost and at competitive pricing people would be losing money, you get in cutthroat competition. That's one possible outcome. Or the other is there's one survivor and you get a monopoly. And so this is the public utility justification for regulation.

that we'll allow one person to produce the output at least cost. At the same time we'll do something about putting a cap on prices so we don't get uh exploited exploited and have excessive prices. Yeah. So that's that's the classic uh justification for regulation is natural monopoly. And obviously though we have regulations in cases where they're certainly not natural monopolies.

Then certainly a big one is externalities. Either consumption externalities. So an example just for a consumer is auto congestion. You're driving uh around and you're in peak period and you delay other people.

Um and you don't have to account for that in your decisions you're imposing a social cost on others in in in in the delay. And there should be something in the market that makes you take account of that decision, but there is no no way the market has is being able to respond that way explicitly. Partly because no one owns the roads, but well the government of course has then has to be a good thing.

uh for for this kind of behavior. And then production externalities, of course, classic one is pollution. Um uh where a firm is producing something and you know, either dump something in water and and obviously makes makes the water polluted and

prevents people from drinking it or gets them sick or whatever, or air pollution and obviously the the bigger ticket ticket concern now about climate change. And again, question is, you know, what's the market doing explicitly to force firms to take account of the cost that they're imposing on a broader uh segment of society.

So the production exercise. And then information failure got a lot of attention to. Um you know, people do not are not acting with full information for for a variety of reasons or even being deceived. And so there there the utility that they're expecting they receive from their goods does not turn out to be what they wind up getting. And so we have a variety of information policies that uh you know try to try to deal with that problem.

Of course the Austrian perspective is to suggest that it's a little unrealistic to To have perfect information be the the goal. The world is inherently it's like saying, you know, friction is a physical externality, a physical imperfection. We need to get rid of friction, but of course

It's hard to have perfect you can't have perfect information. Sure. Sort of a straw man critique of officers. Yeah, yeah. And let me let me just mention the quickly the final one is what I could call public production, that is you have certain services that are socially desirable, but they may not be privately profitable. Uh it may be again it's somewhat related to the scale economies you might see a natural monopoly could be related

problem but one of them simply maybe capital requirements. It's just very hard for a one firm to produce the interstate highway system. Trevor Burrus, Jr.: Orphan drugs, I guess, would be an example potentially of that. Trevor Burrus, Jr. Orphan drugs would be an example. be the be the provider of these goods or services because

even though they might they may even lose money, overall utility is enhanced because the production of these goods leads to social benefits that exceed whatever subsidies the public is going to provide. So these are the textbooks arguments made that that markets need help. And then the textbook says so government needs to a Set the right tax, set the right subsidy, regulate. Produce themselves. Right. And

What I find interesting about your uh book is that instead of asking, you know, whether the textbook's right, uh whether these things are quote failures or not, you say, Well, let's see how government actually behaves when brought in to fix these things. Exactly. So, you know, th there's one thing theory and um certainly that's important in in in framing our analysis of of of social problems.

But then we obviously need to know how the world actually works in terms of evidence. And so the question really that I was concerned with is okay, you know, it's very easy to theoretically point out these problems with markets and and as you said, you know, some of these things may be sort of straw man and you could argue whether really these are

plausible expectations to have, but in any case, you know, who actually per performs better? You know, do do markets make effort to try to correct these failures and in fact

uh the extent of them is is greatly overstated or does government come in and they actually can do a good job and and so the the government intervention is justified or they come in and they make things worse. Collectively what do we really know? And I think the Th one of the things that was really driving me to do this is'cause so many of these issues come up time and time again.

But at least the the public policy debate seems to always start from square one. Don't start from, you know, what do we really know about this? And how can the accumulated body of knowledge really guide us? on how we might aud uh how we might proceed. So, you know, th th this book was really sort of in a sense a a first cut. They're trying to start to assemble what we know.

as as something that we can build off of and learn from and start accumulating evidence and start saying, Okay, look, let's not just treat these things as if they've never happened before Uh let's build on what we know and go from there. And the books are available online and we'll put a link up to it. You can read it without charge. It's uh it's very nice. Exactly.

So what what do we know? Well and how well we know is that that government interventions have turned out remarkably to be disappointing, if you will, across the board. I in it is certainly the case that markets do fail in the in terms of the textbook definition. We do have something called pollution, we do have something call it congestion. Um

Yeah, th but there are other cases where it's pretty hard to find evidence for the fact that really there's very significant market failure, certainly the kind that would justify government intervention. And more troubling is really the lack of evidence of government interventions, whether justified or not, significantly turning things around.

Um now and I'm not even saying that we need to use the benchmark of optimal government policy, just as they are they clearly you know producing benefits. And it was very hard for me to find evidence across the board Um that that led to that conclusion. And let me stress a couple of things. First of all, you know, I focused o not exclusively but almost exclusively on academic research. I really wasn't going to focus on

you know, government reports where they evaluated themselves. For the most part, you know, those there's not many of those. But that really wasn't my purpose. It really w really was more of a scientific assessment of what the scholarly community is is has to say about this in terms of empirical work. Secondly, again let me stress Oftentimes the public approaches these things as as as ideological.

uh that you can predict based on one's political precision where they're gonna come out. Yeah. I think it it's it's very fair to say that I had a good representative sample across the board in in terms of what might think of political persuasions. You know, I didn't just l limit the search to particular scholars. Yeah, th this is a this really is I think an honest reflection of of the economics profession writ large uh what with what they've been coming up with.

So I think it's a it's it's it's a fairly objective story. Now the the big qualification is I've also uncovered there's a lot we don't know. Um th there's certainly room for more work and a lot of these conclusions could could be overturned. But at this stage of the game it it does present a very disturbing picture.

for the eff efficacy of government intervention and a very big warning sign for those who immediately say, let's have the government come in and try to take care of this problem if they expect to have uh very positive results uh to show forward. Of course, listeners know that uh two things about me. One is I love your conclusion, but

Uh but the second thing listeners know is I've become increasingly skeptical about the ability of sci of economists to act in scientific ways in evaluating empirical work. So I'm gonna have a tough time with this. Um I'm gonna have a a um A bit s a bit of a schizophrenic approach to this to this interview clip, so I'm warning you. Um unless the yeah unless we're only talking about publication bias, which is n not a not an unreasonable concern.

I worry about group think, but usually there's a lot of there's potential. But well it also for me, and given my increasing skepticism about the quality of economics empirical work. To me i it it depends a lot and I will press you on this occasionally It depends on the nature of the empirical work. Um I think that's right. If it requires a lot of sophisticated statistical techniques with v lots of assumptions along the way to simplify things, that to me is different than

Averages and and facts and things that are that are relatively transparent. That's right. No, I think that's right. There's gotta be reality checks. And you know, I I tend to try to put some of those in the book with just simple descriptive Amen. data, you know, just you know, what what do we know about the extent of the problem to begin with? Some cases one can do that to see, you know, is there e even something measurable or observable to begin with. Other cases it it it's much harder.

Um I I would also add though that that certainly you know, economists are opportunistic in the sense if if they have a chance to sort of come up with an article that You know, it says all the previous research is wrong, but they're gonna do it.

So you know, I I have to admit I was a bit surprised by the the the consensus that that appears to emerge. I I I want to stress that there obviously were were certainly some frictions in various ways and there's Yeah, there's actually already been an exchange that I've had with somebody on on the antitrust issues.

But but by and large, i it was surprising given the fact that you know we supposedly disagree all the time how much consensus there really was. Yeah. Okay. So let's turn to the evidence. Uh and let's look at a few broad categories. And of course interested readers can uh turn to the book and make sure that the first time.

you can make your own assessment of of whether you find the evidence convincing or not. Uh rather than delve into each study, which of course we don't have time to do here, I'd like to sort of do a I'd like to do a broad strokes Sure. And then talk about some of the the more uh philosophical issues that that the results raise, uh regardless of their of their uh how strong they are.

So let's turn to antitrust and um Yeah, and I think th in antitrust w what was really interesting there was more the absence of sort of positive evidence. Um, you know, w i i it wasn't so much that that I found things that said, you know, government uh investigations of of collusion cases per se, you know, led to significant price increases. I just didn't find evidence of significant price decreases.

I didn't see you c evidence of of cases on anti monopoly cases, certainly Microsoft being the most recent one. where you really could sort of point to benefits to consumers. I think in retrospect people now are sort of saying, what was the point? Of of the Microsoft case. They could have said the same thing. What was the point of the IBM case? What was the point of the IBM case? I mean that's right. Exactly. And now we have and and just quickly make a point that

And now we see the government's investigating Intel. And and and rat I'm not going to comment on that case when I d I'm not an expert by any means on any of that. But i I think what what I'm trying to get at here is here we have accumulated evidence of just sort of the lack of of consumer benefit. from government prosecution of anti competitive beh behavior and monopoly cases, which I think should give some pause before going out and and having another big case.

Yet, you know, we obviously are going to have one here in in certainly one area where you know, chips where we think that the real price of those things have been going down. It's bizarre. It's bizarre. So yeah, th th th this I think is sort of a very topical example of of the kinds of sort of uh tension I see between you know what we seem to be learning from the scholarly assessments and what goes on in practice. So you f when you summarize

And look over the uh studies of antitrust collusion stud studies of trying to prevent collusion where governments intervened. Uh you don't find any consumer benefit. In fact there's some evidence there's consumer harm. Yeah. Um wouldn't the interventionist which I'm not, but I play one on this podcast from time to time, wouldn't the interventionists Claim that well that's just because those are the cases

that you observe. Right. That's right. And it could be that without this occasional foray into threatening action against colluders and and mergers and other things, that There's a hidden benefit we just can't see. Right. The I think the a and it's a and it's a good response. The defense is, well, you don't see really the main point of of antitrust, which is deterrence.

Um, that that without this policy, you know, all hell would break loose and, you know, prices would be through the roof and, you know, collusion would be the rule of the day and and every f you know, industry would have mergers'til we get monopoly and so on and so forth. Now there there has been efforts to try to even address that question. Obviously the difficulty there is what's the control

uh semi controlled experiments that that we run here. And the efforts really have been made on international comparisons. I mean that's sort of the best that that people have been trying to do is just compare differences and regulatory. US and Canada, US and United Kingdom, and that's sort of been the control. of all this and you know, even those studies really just don't see uh, you know, anything that we seem to be doing that that that is all that productive, if you will, and

and producing uh a a more competitive economy with antitrust laws. But I think that's that's still an open question. I think, you know, one of the conclusions is certainly the lack of benefits, but I I don't think anybody can reasonably read this and say we really have a pretty deep empirical understanding of the effects of antitrust laws. It's far from it. I think it exposes

gaps in in in our knowledge here. But I let me stress, I think one thing that that I've been frustrated after writing this It's just the the lack of engagement with economists on this issue, that people still have very strong positions. on antitrust policy, regardless of the evidence, a lot of things are written uh, either pro or con about quote, you know, the Bush administration's treatment of antitrust and in very few of these things, if if any of them, do I ever see

any reconciliation with what the evidence is telling us or what we know. Yet people have very strong air uh beliefs. I think this is one area of economics. Where I honestly think that that it it seems to be pervaded by religion. I think there's more than one, but but that's definitely I would go with um Another small area of of economics where that's true, which is macroeconomics, where there seems to be a rather strong prior set of beliefs that informs people's reading of the evidence.

Right, but w w at least in macro there there is a a a somewhat concerted effort to generate evidence. That's true. And and I think what what is disturbing about antitrust is there just it doesn't even seem to be much much concern about it. And um either mentioning of it. I mean i again I think it the fact that we we wrote this

and pointed this out, i I I just thought it would would strike a chord and it it it really hasn't. Um And people say, well, it's difficult diff uh difficult to get data but I think even the agencies who have economists and know about this, you know, can make things easier. Uh and and in fact I noted that one of the responses to this work was a task force uh with one of the antitrust authorities to say that they were gonna do something like this. But you know, then they dropped it. So

I think in in many ways this is this this this was an an exploration that was as disappointing for policy as it was really for the economics profession. Well we have to confess, well maybe not individually, but certainly collectively that um whatever that means. Uh that economists have a a large personal stake.

financially in the antitrust wars. That's right. That's a lot of people say they they consult and that's really what's driving extremely lucrative. Yeah. A lot of this. At the same time, you know, the evidence is what it is and Uh again I think if if there were compelling benefits, uh it w it would be good to see them. You know, what what also is the the more open question, just just getting to your macro, micro, macro uh discussion we'll have a a bit later is the lack of a broad

overview of the you know, the over the social deadweight loss for monopoly in the United States. This was something that Harburger did a long time ago and admittedly it was very crude.

and, you know, lots of assumptions and all that kind of thing. But, you know, it suggested a small one and it it is an important measure I think for us to to try to work on, which is something that is does not go on industrial organization, is some sort of overview about how industries are performing and subsequent uh calculations of of what Hartberger was doing. We really haven't haven't exposed again that one of the big problems in the US economy

is high markups because of monopoly power. So I think that that's sort of the kind of reality check that that would also be very powerful to pursue. But it it it I think again one of the weaknesses of of of IO is is is just not really giving us

enough hands-on sort of empirical overview of the performance of industries. We know though it does not seem to be uh troubling. IO being industrial organization. Yeah. Um So one one I think defense of the interventionist approach would be Well, okay, the U.S. economy looks awfully competitive and in in fact I'm a very Schumpeterian uh economist on this myself, uh, when I'm in my own shoes, I you know, I I see tons of competition everywhere. I I see monopolists who are supposedly getting

Suppose monopolists whose share is growing suddenly see the market disappear because of a competitor that comes out of the blue. So I'm a I'm a big believer in in the power of competition, but the s the contrarian view says, well, it's

Sure, there's not much monopoly because that's because of the eff effective threat, this deterrent we talked about before. And what would be the the big gain from throwing away this apparatus? So Being a a non interventionist, I'd say, Well, let's get rid of all this antitrust stuff. And the contrary in view could say, well, it's not that expensive and as a result, maybe that deterrent thing is important. What do you think of that? Right.

Well, I guess th th there's two things. I uh at at the t for the time being if if one w one dr wants to immediately draw policy applications of what we're saying It's that you know, one should focus at least on the most egregious cases. You know, merger to monopoly, you know, there may be two firms in the industry, you allow monopoly together. Yeah, that's something arguably, you know, y you you certainly don't want to s uh want to allow to go through.

Um but there those are certainly rare cases. I don't let them merge. It's one of those issues where once you how do you define the market? But I have no problem with Ford and GM merging. Uh and then the the s um uh the the uh the second issue, you know, o obviously blatant collusion is is something that that we certainly don't want to be encouraging.

But now, you know, your your comment about um well what's the harm? Well the harm obviously is how the how the how the policy and the political economy of all this is gamed. that is uh arguably Microsoft's competitors were the ones that that had a lot to do with bringing this case. Uh there are obviously huge transactions costs. This requires the attention of management. for a considerable amount of time. Uh a lot of resources go w go into that. Merely the government isn't spending

uh billions of dollars. But you know, th this this can create a culture in which you get a mindset where your managers spend more time worrying about gaming the government than they do producing better products. I think it's even more explicit in in regulated industries and even the deregulated industries, uh that they grow up their whole life

with the government and it's very hard for them to sort of cut the court so to speak. And and I and I think there are costs to that. Very very big cost. No, I do too. And and and it can be very difficult to measure, but it's not the climate that that that one really wants. where one can think that, well, you know, I'm I'm having a problem with a competitor and it's time for me now to start, you know, lobbying my congressman

You know, to to to then turn to the antitrust authorities and and help solve my problem. Which arguably is what's going on with Intel. Yeah, no, I'm uh I'm with you there. The uh it's an example of uh the destructiveness of rent seeking. And uh we did a podcast with Mike Bunger on that and we did one with Don Boudreau on this issue of antitrust. We'll put links up to those to remind people to check'em out if they want.

And I think it's a pervasive problem, you know in in all areas where government is involved, let's s let's face it, that it's going to be political and people are always going to try to use this to their advantage. Do we have any measures of the magnitude of that cost?

No. That would be a good thing to look at really. I think that would be something obviously th that would that would be good because I I think it I think in i particularly actually in the industries I know better th a lot of the recently deregulated industries. be fun to do a case study at at a corporate level of how many hours of the day or of the week the CEO spends on these issues and how much how big that legal staff is, how many Right. Yeah, it's foregone productivity of the economy.

So let's let's move on though to a different area. Let's move on to safety and Safety regulation, other forms of of what you call social regulation. All right. So so so so social social economic regulation um you know deals you know primarily with with the externalities, um

uh safety cer certainly being one of them. Um and and workplace safety, I think that that that's that's a good one because you actually have some of the descriptive evidence of your workplace accidents going down But then when you try to isolate what the effect of OSHA interventions are. So um that that's the regulatory agency that's supposed to be inspecting

you know, safety of plants, you w people tend not to find a statistically significant effect of the OSHA regulation. So in in the world it seems like things are getting better. And this is the value then of regression analysis for trying to isolate the counterfactual, that is, really what can cut what can we attribute to to the policy as opposed to other factors.

And and then I think also what what's helpful in in this area is then more detailed study of and of how these regulations actually work in practice. And lem let me just stress that point. I think Yeah, one of the att one of the one of the attractions of a very eloquent speaker l like the president is that one gets the idea that it is someone of of his intelligence is actually sort of carrying out and performing these regulations.

Yeah, w when in practice, you know, you know, you're you're not getting the president to go to inspect plants, you're not g having the president go inspect toys. You're not having the president inspect drugs. You're disillusioning me, Cliff. You have people that you know who are who are uh in these regulatory agencies Yeah, it's it's it's not uh particularly well staffed in some cases. Uh they're very spread out as as we learned on the

on the uh toy testing we found out there was like one toy tester for for all these toys that were covered in for China. No wonder a few of them got by that weren't particularly safe. And OSHA is a similar type of thing. And I think between just the general morale and the difficulty of sometimes trying to persist with with some of these uh investigations Yeah, we really don't get an agency that that is particularly efficacious at sort of identifying problems.

Um, this is not to say that some of them may not exist. It just again, I think just the nature of how these things are done is very cursory and we don't really see much. Now fortunately the market in this case provides huge incentives. for firms to have safe workplaces and to produce safe products because obviously they could be heard in the courts for for we have a liability system and of course killed in the market uh if um

uh things are unsafe. And of course even in the labor market. If if you have very unsafe workplace conditions you have to offer a compensating differential.

What about imperfect information? The workers don't know that, so they get exploited and taken advantage of and they don't get the higher wage'cause they don't Well first of all i i think in the area of safety that that's one area where it's pretty clear when you know an airplane crashes or when a mine caves in or when you know people repeatedly have their hand chopped off from a a a machine that That does it work well. And I assure you in in increasingly

uh the the word gets out about information at work places. Um you know, it's interesting, at least in the economics professions, I'm seeing for the for the first time sort of These job market blogs with just an awful lot of information going back and forth about potential employers. Uh I can only imagine that that's got to exist in in other industries as well. So it it's pretty hard these days to sort of hide the fact that that working f for a particular firm is

is safe when in fact it's unsafe. So again, these are the kinds of market forces that that have gotten even stronger. Um and then when you look at the the realities of the the regulatory agencies in terms of you know how they're run, the constraints on people, the people themselves. Uh you know, it it's not surprising that that we really don't find much evidence that these things are particularly effective. So just say one word about the your your um comment on regression analysis.

It it's always tricky, and this is uh an issue we've talked about before, it's always tricky to quote, hold other things constant. The world's complex. What I would say is interesting about the safety issue is that when I've looked at those data It's pretty clear, it it's overwhelmingly clear that the overall long term trend in safety across every dimension of our lives, auto safety, workplace safety.

It is getting better and better and it's been going on for a hundred years, fifty years. As long as we have the data for it, that's how long it's getting better. The standard economic explanation I think would be that safety is a normal good, that is The jargon is it's a good that's positively correlated with with income and prosperity. As we get more prosperous, we're willing to devote more resources and to these issues and firms compete in providing safer and safer products.

Because as we get richer consumers are willing to pay a premium for safer products or To take a safer job, et cetera. Right. Because the quality of life is is getting That's better. And that's what this is actually about. It's it's really raising the quality of life. So when you look at the data and you see that the trend is unchanged by the creation of OSHA.

That doesn't prove that OSHA didn't make a difference. It's possible, and this is of course a a real possibility, that it would have been worse without them, that the trend would have been different. But the key point to me is that puts the burden of proof on the OSHA defenders. Not to just say, well, look, the law says it's more safety, but the data say no improvement. So you show me.

OSHA defenders or uh auto safety defenders, that that the mandates as opposed to the bottom up emergent improvements that have occurred, that the top down mandates were the things that that that they that they

Avoided something that would have been worse. And you have to you have to make it a lot of things. But also mechanism. I think I think also what what we w and increasingly we want to get into is a much sort of a deeper micro understanding, you know, m just even abstracting the the data away from the data. Um and just say behaviorally, what exactly are these inspectors doing um that's that's telling the plant

you know, how to operate more safely. I I make this point on airlines all the time. You know, what exactly is it the FAA is going to tell you know aircraft manufacturers and airlines that they don't already know. If anything, they've they've taught these people something. Um you know, we w we see the Dreamliner has has just been uh

tested by Boeing, but it now has to be certified. I I couldn't help chuckle and say, no, wait, the FAA is going to tell Boeing whether they think their plane is safe enough to fly. Yeah. Right? I mean, you you imagine the quality of the people who work for Boeing on that plane and the kind of training and knowledge they have.

Oh the textbook, the textbook. So I I think those are the those are the kinds of things that that that we need to press to get sort of a deeper understanding. Now, you know, there there can be mechanisms um you know, w where could where one might be able to point something out. But I think that's very important to do and and just just jumping ahead, I think that's also critical.

in in the area of obviously financial regulations and and these things. Again, I'm no expert, but what I'm looking for in any sort of s you know, proposed legislation is tell me exactly the mechanism, you know, by which you really are gonna reduce risk without, you know

truly throwing out major incentives and innovations and these kinds of things. I think, you know, one has to sort of go beyond just the abstraction, there's a problem out there, we need the government to do something, and really point specifically what exactly it is Well the Boeing example is a great example because It really highlights to me this distinction you're making between market failure and government failure. Now it's true that Boeing

really isn't eager to crash its planes. I it's not good for Boeing when the plane crashes. Not good. Not good. But you could argue and and people do all the time. Oh it's true they don't want a the plane to crash, but they're gonna cut a corner here and there'cause they're motivated by profit.

And the the economist response to that is, okay, true, but they care about profits tomorrow and profits today to the extent it's a organization that is ongoing with repeated interactions, its incentives are going to be Probably very closely aligned with consumers. And so the counterpoint though is yeah, but not perfectly. And so what the government can do, emphasis on can, can do is just to make sure that they haven't cut a corner here and there to boost, say, third quarter profits.

uh because of some stock crisis. Okay, fine. Good point. But then the question would be your question, which is I think Is is is so phenomenally appropriate, which is fair enough. Now tell me what the actual structure of the FAA would be, not what you imagine it could be, but tell me how it in practice reduces that corner cutting as opposed to just making life more difficult

For Boeing. Well also and pointing out pointing out to Boeing that it is cutting a corner that Boeing itself doesn't realize it's it's doing and hasn't thought carefully about the trade offs. Because I think again, it it is clear sometimes in some of these investigations with the FAA that the airlines realize, okay, yeah, we're not we're not ex we're not we have not changed our configuration of how we have to put the the the screw on to, you know

some some little part of the plane and they say, Look, we're still gonna fly these things. We know th this isn't in line with you know, procedure, but the fact of the matter is there's no way that this thing is gonna compromise the flight of the plane.

And they know these kinds of things. I mean, i i when when they cut corners and y th they have serious repercussions, you know, it it's not just also profitability, you know These can be these can have, you know, negligence issues that that can also have criminal uh implications to them. And a lot of these people do know this. Now this doesn't mean they some of them still won't take a chance, but this this becomes very, very serious, life changing things.

that just go beyond hiding behind a firm losing money. I mean human beings can can go to go to prison for this kind of stuff. And uh this is not something also that that people who self select to go into this line of work want to do. I wouldn't sleep so well at night either after the fact. I th I one would hope also. I mean and I think it is true. A lot of these people have had very serious problems when

Uh when sa when bad things have happened. And again, you look at the observed data, things have in the air area has gotten remarkably safe. Um let's talk about the I I just I just want to throw in one other thing if I could. That you one thing I also want to stress and and this is uh a a point that um

uh Sam Peltzman, who I know has has has uh contributed a podcast before, has made is just sort of market robustness. You know, how how people themselves, the markets themselves, try to respond to externalities. So, you know, one also sees this as as a way of at least reducing the costs of them. And to be specific, you know, a nice example is congestion. And w w what we see in terms of a market response is well

Obviously the congestion that you yourself experience is something that's under your control. That is in the the buzzwords of economics, it's endogenous. And the way it's under your control is where you choose to live and where you choose to work.

And quite understandably, if someone really dislikes driving um in congested traffic, they're going to make location uh residential location and employment choices that try to reduce that congestion costs, albeit they may be trading it off and paying a li a little bit more to live closer in uh in in their housing. And so y we s often will see data in terms of estimates of value of time that people who have the highest value of time tend to have shorter commutes.

And people with with low value of times have longer commute. And in a sense that's the market's way of trying to respond to an externality. Uh we the sort of a classic example of that um is in the area of of of noise, you know, people who don't like to live near runways or airports because they don't they have a low tolerance for noise. And I I remember one response that I that I heard to that as well. In Boston they've located the school of the deaf Right near the Logan Runway.

So'cause the land is cheap and then it doesn't matter. And there you have people, you know, who obviously are are not going to uh uh be annoyed by the noise as much. And so these are these are then interesting ways uh that we learn how markets uh try to respond to these things. Not fully, but You know, th this also can can be an important consideration. And I think again in the information area that's also we're we're we're seeing. You know, I do not expect

in the financial area that people make the same types of mistakes they made this time around. There'll be different ones. But but there's going to be learning. Yeah, no doubt. Um

Let's turn to what I consider the classic case where government regulation uh may have improved matters, which would be pollution. Um Now there is some private incentive to reduce pollution. Pollution is is is foregone uh efficiency. It there's a natural long term trend toward uh less waste in in all c aspects of production because it's waste is costly. But there's still some and there's always a temptation to dump that waste. into somebody else's air, somebody else's water. So how what's been

the net effect of um government pollution regulation. So the the the the Uh one of the things that that I also just want to make clear is there are alternative ways in which we address uh pollution or externalities. We can use the price mechanism. uh in which you you you you try to put a charge uh for somebody for for doing this or a quantity mechanism by saying, look, there's there's certain technologies you have to use

to abate the pollution. The the government's preferred approach in these things is usually what they call command and control. And what happens is is that we can see, you know, sort of almost by force air pollution redu uh going down and automobile pollution also has gone down. But again, I think there's been a very heavy hand

uh by the government here in terms of pushing things too far back. In other words, you know, at each point you can look at your output and see is your social cost exceeding your social benefit? And I think what what the evidence has shown is yes, we we have had benefits from the reduced pollution, but they've come at very high cost in terms of in in the case of Old Beal

increasing the costs of of cars, um, in in terms of production, you know, inc increasing cost to firms. So what most of the net uh estimates appear to be is that it's fairly balanced, if you will, that what gains we have made from pollution have have often been pretty much balanced out by the higher production costs. Uh

of doing so. And I think again, this is the kind of then lessons and concerns that we have now with the current climate debate that You know, are we gonna go ahead with a policy While certainly motivated in in a time of great concern, is too heavy handed and that we're gonna have to basically sacrifice far too much of our GDP. to make the improvements uh in climate change that are determined. Do you want and and I think again that's that's an important lesson that Yeah, one point.

just generally to the outcome that we have a goal, we want to reduce pollution and we can do this and we're not going to worry about the cost. But, you know, we ha again, we have limited resources and if at the same time, you know, we want to reduce pollution but then, you know, spend money on uh subsidizing somebody else for something.

uh in in the social area, we're not gonna have the resources to do that. And so this is sort of the classic problem that we often have uh with with a lot of these sort of social regulations. We have this goal We want to reduce we want to achieve a particular level uh of the activity and we really pay very little attention to the cost. And I think that's sort of what's come out of the the evidence on uh on um pollution.

So you're suggesting that we could have gotten those same benefits But at a much cheaper cost. W what mechanisms could we have used in the case of, say, auto pollution of the air that that would have been cheaper than what we actually did? Uh explicitly pricing emissions. Uh believe it or not, these days we could actually monitor um, you know, the social cost, if you will.

uh pollution and there are ways of uh in doing this without sort of intruding on the uh you know where the driver is going and all these kinds of things. And so you can you could uh you could have actual price Uh is is the recommended approach or tax some tax uh tax for Imi emissions would be certainly a way to do it. Uh a similar type of example also occurs with airplane noise. They have command and control regulations that they have to make airplane engines that have certain decibel level and

They just have to meet these standards. Now similarly you could have a quote noise tax. to airplanes and say, look, you know, if you're gonna go over this, this is what you've gotta pay. And again, it it's similar to pollution. If you actually look at the descriptive data, you'll see the decibels exposure, if you will, to to neighborhoods around airports has gone down dramatically.

But it's the cost then to the plane that this has now increased the capital cost to the aircraft manufacturer. It r um made their planes uh obsolete so to speak and reduce the value of their capital stock. These things obviously are more abstract and not the kinds of things that that regular people worry about, but this is going to be reflected somehow in in in their prices.

Um and these are the kinds of trade offs that we often make. And I think that there is in a sense that the general characterization, sort of an engineering mentality that often comes about with a lot of these policies in that we just you know want to achieve you know, the the technological goal and the cost becomes secondary, obviously this spills over into things like public production. So we want to build the extension out to Dulles Airport on the Metro line.

And I'm sure when it's all done everyone will be excited about it. But loss beneath this is gonna be the billions of dollars and cost overruns that's gonna go on in the production of this uh extension. But again, those kinds of cost benefit trade offs generally are not f uh foremost in public policy formation. Other people. Not the residents who will use the system.

uh distribution, if you will, of costs incurred by the taxpayers, you know, i enables these sort of policies to move forward with With the cost being spread out because indiv you know, people say, Well, you know, it's not costing me very much or not enough that I could care about and you have oftentimes localized localized beneficiaries. Certainly in the in the case of the pollution that we talked about, yes, that we all benefit, but again we're also all paying.

Yeah. Let me quibble a little bit, maybe a lot, with your um with your tax idea. It it seems that it's better to have a tax uh than the uh command and control mandate of a particular technology, catalytic converter being another example of where instead of just saying uh there's a

For every amount of pollution of of say sulfur dioxide that the car puts out, or whatever cars do nitr whatever the bad things they put out, I don't remember even now. Whatever bad stuff they put out, for every amount above a certain amount, or every amount that you do produce at all, you pay a tax.

Instead they said this is the technology you have to use and of course I think we talked about this on the program before. At the time, Honda met the same standard of pollution that the catalytic converter produced in other cars, but they were required to have one anyway. So that that was an obvious example of of a distortion that was for political reasons. But my my quibble and I actually it's more than a quibble is

You said we could we could put a tax on because we we could measure the social cost. What we can measure is the emissions. The connect now, that's the technological improvement. It's still a challenge to figure out what the social cost of those or the harm that those emissions cause. Sure. And there's still going to be political pressure.

To set the level of that tax uh in ways that are not necessarily good. Trevor Burrus No, that's right. That's right. I mean obviously the role of the economist is certainly to inform uh those measures and this is exactly what also goes on in in setting congestion tolls. Yep. Uh you know, again, you know, the idea is to use a

you know, estimated value of time and this is what the e economists you know contribute to this debate, uh or c contribute to the to this policy, if you will. But no, it's true. As a as a practical matter, um th these things can become politicized very quickly and even though you know uh economic research will have particular bounds of what these pollution costs are

uh this doesn't mean that in practice that's that's what we're gonna wind up getting. Uh actually we in we're even seeing that already in uh in a highway example, I was just saying that the inter the uh intercounty connector uh tolls of toll schedule has been announced and they seem very high um to users or at least to to observers if you will. And I'm sure there will be political pressure to probably reduce those. Yeah.

Um well I'd like to do a show sometime just on traffic congestion'cause I think the uh the politics and the economics that are quite are quite interesting. But we're getting low on time. I wanna turn to a false a couple of philosophical questions here at the end.

Um So let's accept your Your findings and they're not yours per se, they're the a summary of what many, many people have studied, which is a a sort of a gloomy picture, uh, for some, that that government intervention often doesn't achieve what it was

uh intended to do, often serves counter goals counter to what that was intended to do, or at least described as doing, and certainly counter to what textbooks suggest would be the case for government intervention. And I guess we could think about two broad Reasons for that, if taking it as true.

One would be just incompetence. The incentives aren't there for government to do the right thing. You talked about morale. There's all kinds of there's the world's competition. Technical challenges. Technical challenges. The second is of course more sinister. Uh and that would be the idea that economists have been writing about for a long time and political scientists as well, which is what goes under the broad name of regulatory capture. That once a regulatory mechanism gets set up

The people who are regulated have the biggest incentive to make sure it serves them rather than a so-called public interest. Any thoughts on which of those is more likely to be the case? Do we do we know anything about it? I think it's the latter. I I think the political economy is is really what's driving an awful lot of these things. Um and and it's just just interest groups in general. Um, you know, a lot of times it's it's not sort of capture per se, but

you know,'cause there's not somebody being regulated. It's just, you know, people who, you know, have an interest, are well organized. America is increasingly getting better at you know, lobbying and putting more money into it and the returns from it are are really high. And I and I think you know, for example, in the highway area, you know, y you'll see particular users, you'll lobby for certain roads. It's not that they're being regulated, that they want the government expenditures.

Um but obviously in in cases of regulation you can see that they they want to use it you know, to uh get protection or you know, hurt their competitor or what have you. And I think the answer to the to the question is extremely important because it then leads to okay, where do we go from here? In other words, what's the big policy conclusion? that I get from all this. And and obviously the the policy conclusion has to derive from, okay, what is the underlying problem?

And I think the underlying problem is is our interest groups and political economy. And I think that the that leads us to an answer that really calls not for the wishful thinking of improved government policy, but I think more privatization and at a time when markets may be suspect

more reliance on them to solve these kinds of problems. Now obviously, you know, that's gonna take time. But I think one thing that that I do wanna point out is Yeah, we we didn't talk much about natural monopoly or regulation and I think the reason is because the major examples of those cases are ones where we ha now have deregulation or at least partial deregulation. And I think that what we've learned is that we have been able to get gains when we withdraw government interventions

to try try the try to solve these so called market failure problems and rely more on the market. So my sense is that for a lot of these things, given the power of a lot of these interest groups, that, you know, where where we try to get market mechanisms and particular privatizations, we're gonna have a much better chance of solving these problems. And and some of this stuff is is, believe it or not, on the horizon.

Wouldn't the uh skeptic on the other side say that the Enron story in California's failure shows the dangers of of deregulation? Well these natural u natural monopoly utility markets. Well what's interesting is remember you it was in the sense it was the market that that sort of outed Enron. There was nothing Yeah, I know. Yeah, there was nothing about the government that outed them.

I'm talking about the energy part of it, not their frauds and I'm talking about the The attempts of California Enron was sort of t you know involved in this because in terms of how the well Yeah, we have a an example where something is called deregulation that that was so mismanaged and sort of went went in the face of what what we really were trying to accomplish there. I think that that that that is a very important caution.

on on how you're you're you're going to withdraw government intervention, if anything. That was a big setback for that for that thing. But it was it was mismanaged uh terribly. I I I would add that in in the things where I see there's chances for privatization. airports, metro system and even highways, that it's very important to run very carefully designed experiments so we avoid the California energy problem. What kind of experiments do you have in mind?

Well I you know we believe it or not, we actually have some of these things on the books. So, you know, th there are airports now that are applying to you know, try to become private airports and and Chicago Midway was was the first major one. where where that would happen. Uh unfortunately they've they're during the crisis the their funder wasn't able to come up with the money but they still still hope to do it.

But really it would be the idea of of transferring ownership, complete ownership of these facilities in private hands and and and let them compete and see what happens. Seems like a good idea. You can imagine, you know, a bus service in in our area. Just completely contract that out or let let a private provider take it.

And get get government out and and and let them let them make a go of it. Yeah, I recommend the podcast of Mike Munger on the Chilean bus system that went from the chaotic private system to a government run system with Very bad results at the time. Um

Any um any thoughts on the feasibility of the politics? It's it's easy to say We ought to move to more privatization. True, the political case for that is trouble is hard right now because there's uh skepticism about markets, I think misfounded, but it's definitely a reality that a lot of people are worried about.

about private incentives. But it seems to me that the special interests are the bigger threat that your I think correct analysis that special interests steer um Regulation a la Bruce Yandel's bootlegger and Baptist analysis that that you get a uh coalition of people worried about safety, but when the actual safety regulation gets written, it tends to be written by the people with a stake in it and they manage to write it in a way that it doesn't serve the public interest.

How might the public inter how how might the special interests be quieted that would let us get to that privatization outcome that would be much harder for them? They'd have to they'd have to work harder. Well, I think this ties in with your sort of micro macro question. That is What what often we see well not often'cause we don't have these that often, but when you when you see major policy reform, you know, you you really need some political entrepreneur to try to sell this to the public

in in a very effective way that can head off the interest. Let's let's keep in mind that, you know, deregulation of inner city transportation was not sold as you know making trucking more competitive or, you know, airline scheduling more efficient or or whatever. I mean the the details I think of of the policies were less essential than the broader vision of this is a time of inflation during the seventies. You know, here here we think that regulation is elevating prices

Deregulation's gonna increase competition and it's gonna help us fight inflation. Yes. I think it it's it's those kinds of slogans and and broad visions that that I think are much more appealing to the public than often the the details of something that they themselves are not all that particularly in v uh vested in in terms of bigger picture things. Now I think in in a sense now this crisis provides us with an opportunity because you look ahead

And say, you know, what are really going to be the major big big ticket, you know, economic concerns that w that that face us in in the coming decades? Well one is certainly budgetary concerns. Um, you know, although the president doesn't seem to emphasize this, you know, we're gonna have to pay a lot of this back. And governments already at all levels are running huge deficits. Uh at the at the same time we're also going to be looking for areas to spur growth, innovation, what have you.

So when I start talking about privatization let's just say again of transportation which I know a little bit about. You know, I'm thinking okay, here are ways that we can sell this as something that will help budgetary problems facing governments because here are things that they can sell. uh reducing pressures on their budgets, they'll get revenue certainly, but also then they won't be incurring the costs of maintaining and operating these systems.

And so that wa that actually can be quite effective in dealing with that problem. At the same time, we also can turn the private sector loose. in an area where they possibly may come up with some pretty exciting innovations. Um so that's just sort of an example of of how we might want to approach this problem. Think more broadly about where it could help you know many people in an economy and try to sort of sell the more specific policy in that context.

My guest today has been Cliff Winston of the Brookings Institution. Thanks for being part of EconTalk. Thank you. This is Econ Talk, part of the Library of Economics and Liberty. For more Econ Talk, go to econtalk.org, where you can also comment on today's podcast and find links and readings related to today's conversation. Engineer for Econ Talk is Rich Goyette. I'm your host, Russ Roberts. Thanks for listening. Talk to you on