 Welcome to Free Thoughts from Libertarianism.org and the Cato Institute. I'm Aaron Ross Powell, editor of Libertarianism.org and a research fellow here at the Cato Institute. And I'm Trevor Burrus, a research fellow at the Cato Institute Center for Constitutional Studies. Joining us for a record fourth time on Free Thoughts is our colleague Peter Van Dorn. He's a senior fellow at the Cato Institute and editor of Regulation Magazine. Peter, today we wanted to talk about a subject that shows up in free market economics sometimes, particularly when libertarian free market economists talk about things that's called public choice theory. It's a term some of our listeners may have heard, but what is it? Well, let's start with what economics was before public choice. So economics is the study of individual preferences and the interaction of those preferences with firms in market settings. And we teach students under what conditions in effect people's preferences and the firm's ability to produce, interact and supply and demand settings and at what prices, goods would be traded. And we talk about the famous word equilibrium, right, when in effect the market is at rest because every preferences that exists has been satisfied and firms are making normal profits, blah, blah, blah, blah, blah. Collective choice was not traditionally part of economics. And so public choice is simply the examination of the properties of systems of aggregating individual preferences, not into individual choices and results, but collective choices and results. It's that simple. So that makes it just another branch. It's just part of economics, regular economics, but people hadn't really applied it to collective choice beforehand. That is correct. And the part that you described, there are at least two distinct branches of public choice economics. One is often called the Virginia School, which is associated with the scholars Buchanan and Tullock and was at Virginia Tech and later the University of Virginia and now often is associated with George Mason University, both its law school and its economic school. But there's another completely different and not only completely different, but different origins part of public choice theory that I'll call for lack of a better term, a kind of Caltech game theory, very mathematically driven part of public choice, which is about the mathematical properties of voter preferences and different systems of aggregating those preferences into collective choices. And does it focus mainly on voters or does it focus on the behavior too of politicians or is that more of a Virginia School type of thing? In some sense, the Caltech School of Public Choice focuses not on voters or any real work. It focuses on systems, the properties of mathematical ways of aggregating preferences given certain mathematical characteristics of those preferences, which I can describe. And whether or not there are paradoxes or properties of various systems that lead to perverse results. So it's theory. It's idealized voters and idealized governments that don't exist in the real world. And in my training in graduate school, I was mostly raised on that kind actually rather than the so-called Virginia School, which to be honest is more politicized. It's more explicitly anti-government in its origins, whereas the voting theorems that we can talk about today are ideologically neutral. They're just descriptions of if people have these kinds of preferences and we use this mathematical machine to aggregate them into choices, what's likely to happen? You describe public choice theory as being different from the rest of economics in being about collective decisions as opposed to individual decisions. And I'm curious how that relates to another breakdown in economics that may be more familiar, which is between macroeconomics and microeconomics, where macro is about the kind of emergent things that happen when lots of people at the micro level are making these different sorts of decisions. So is public choice a kind of macroeconomics or are we talking about something different? It's a good question. Let me try to give the following. Macroeconomics is the study of whether individual rationality in markets can add up to collectively bad results strictly in markets. In other words, even though everyone's a price taker and everyone does all the things one learns in microeconomics, could all that individual interaction somehow add up to a suboptimal result collectively not in government but within markets where we all could agree that, and again, I think on a previous lecture I've said when anyone ever says we could all agree that, be worried, but I'm about to invoke that. So for example, even though everyone's a price taker and everyone does this and that, is there a collectively bad economic result? That is what macro is about. And then there are theories about, hmm, everyone isn't a price taker. There's sticky wages and sticky price theory, right, and that's a market failure and that's why we need, you know, fiscal policy and monetary policy, et cetera, et cetera. Public choice theory is simply using the tools of economics to study, in effect, collective preference aggregation rather than what macro is, which is whether individual action in markets always adds up to good economic results defined economically. So choice theory is explicitly about, you could think of it as the economics of government, not in a public finance sense but in the optimization notions that one learns in microeconomics applied to thinking about collective choice. And it seems if we're thinking about the first people who really kind of talked about this stuff, which I know there's condorses and people like this going back. 1700s. 1700s, yeah, and then Kenneth Arrow and Anthony Downs and talking about voter behavior but it seems like the way you'd get into this subject is to start thinking, well, sometimes in the world I choose things in a market, for example, when I go to a department store I choose things and things are different prices and I can have intensity of preferences. I can pay a lot for something or I can pay a little for something and only I get to decide this. Everyone doesn't get to decide this but in voting neither of those things are present. Everyone gets one unit of strength for their vote. There's no intensity and apparently everyone can kind of get a choice to decide or not. So this makes it entirely different if you're going to analyze everyone deciding to go to where are we going to eat tonight. One way you can do is like, well, people put their money down and whoever puts the most money down gets to decide where we eat tonight. Another way of doing it would be to vote and both of those would have different incentives that they would operate under and that's sort of I think how they started thinking about this originally, correct? That's correct. That's correct. And Condorcet's theorem was in 1700 and something and it was forgotten. And then a mathematician economist named Duncan Black in 1958 kind of rediscovers this and Black's theorem is about what we now call the median voter theorem which is the simplest theorem in collective choice and I can sort of walk us through that. But it basically says if there is a conflict dimension. Now, you used this term earlier today when we were talking off the record and a conflict dimension sounds like something either an international relations scholar would use or a football coach or something like that or science fiction. People like me always think in terms of an X and Y axis is always in my head. That's why it's sometimes hard for us to communicate. So along the X axis is say that the town budget for plowing snow in Vermont or how much should we spend on defense or name any other classic public expenditure dilemma and that is the conflict dimension. Some people want not much spent on this public sector activity and other people want a lot more spent on this public sector activity and public choice call scholars call that the conflict dimension. So assume we have a conflict dimension and we agree that that's what we're talking about and we have to come to a decision. We have to find out how much to spend on the snow plow budget or how much to spend on the F-15 fighter or any of those kinds of things and make believe people disagree about how much to spend. And here's where we have to imagine a graph and I have to talk about the Y axis. So make believe everyone has, you talked earlier about intensity. So we're down to a vote. But underneath the vote is actually I like this a whole lot more than I like that. But in a voting setting I can't express that. I just have to say I vote for this proposal or that proposal. But one can think of voters liking or disliking how much to spend as a continuous function that has a maximum at some, what public choice scholars call an ideal point, i.e. I want to spend $1.2 million on snow removal this year in my township in Vermont or I want to spend 2.6% of GDP on national defense for the United States. And because it's a dimension, we can go much less spending or much more spending left or right for my ideal point. And then the Y axis reflects a function of how much I like those points to the left of or the right of my ideal spending point. And one can think of that function as probably looking like a normal distribution. So as I get further left or further right from my ideal point, I like that amount of spending less and less and less and less. So as I get further away from my ideal point, the function decays in space down the Y axis. But that decay doesn't have to be a mirror, right? So you could have situations of trying to think of how to make a concrete example of like say national defense where my, if I'm a hawkish sort of person, it might be the case that I think we should be spending $1 trillion on national defense. But I would, well, that's the ideal and I think that probably $2 trillion is not as good as $1 trillion because a lot is going to be wasted. It's I'd much rather spend $2 trillion than half a trillion because not spending enough is far worse than spending too much. So in fact, we're going to get into which is Aaron always asked the good questions. This is happening everywhere. Okay, fine. What about me? You're quite correct. I'm not putting any presumptions on the shape of this preference function or that the right side of it looks like the left side of it. Just that there's an ideal point and that is your maximum and you dislike all other kinds of spending other than ideal point as we go and as we go further away from that ideal point, you like those things less and less and less as we get further away from your ideal point in both directions. But again, going back to Aaron's question, as I was thinking about this on a more macro or micro level like a household, let's say that some five-person household, three kids, two parents uses a voting system for how they're going to spend their budget and so the kids have this, let's say Aaron's daughter has a very high princess desire, right? And she wants to spend a huge amount of the household budget on princess stuff, right? And then show any deviation from that. But it seems also questionable because she probably wants to spend 100% of the budget on princess stuff unless she realized that there were costs, there are correlative costs if you spend too much because you could have this idea that everything above this point is okay. You could never spend too much money on princess unless it had other things you're denying yourself because you're spending too much money on princesses. So when she stopped having a house and food, then she might have less of a preference for princesses, but that's her own sort of irrationality function in all three two. Right. We're assuming a budget constraint. We're assuming, right? We're not introducing any weird tricks like assume a five-year-old and what, right? We don't do that. The political economy of five-year-olds. Right. Yes. Right. The other function I've just described is called the single peaked preference function in a conflict space. Okay. So you have an ideal point and then you like other alternatives in either direction less well than the ideal and, and you never like, there's never an upswit. So as you get further away from your ideal point, you never have an upswing again. They just go down. You like things less and less and less forever as you get away from your ideal point. And so then we aggregate a bunch of people with all those different functions together. Condorcet's theorem, right? The median voter theorem simply says the following. If the axis of conflict is agreed upon and- Can you clarify exactly what that means? That we're talking about the, so- We're talking about something. One thing, yeah. And we agree that's what we're talking about. We just disagree about where, where we ought to be within that. I'll bring this up later when we talk about germanness amendments in the Senate and the House. And this gets very important. Yep. So I have to start out at the beginning though with the, the simple median voter theorem, which is if we agree on an axis of conflict and we array all our preferences within that conflict space and every voter, every legislator or voter has a single peaked preference function, then a sequence of majority rule choices over everyone's ideal points where we vote on things two at a time, one, one proposal versus another, then a sequence of majority rule decisions will result in an equilibrium at the preferences of the median voter. Okay. Can we break that down into English? Right. It makes sense, but break it down into examples so you have a bunch of people with different preferences. So there's the three of us, right? Yeah. You have five voters or three. So Aaron wants to spend blah amount on snow plows or, or the fighter plane and you want something other than that and I want something different. So if I want to spend a million, Trevor wants to spend a million five, you want to spend two million and, but we never, we never say the voting system isn't such that we say everyone write down how much you want to vote, you want to spend and we'll average them or whatever we instead say. No, we just have a proposal for, so the floor is open, right? We have. I propose 1.2 million spending, yes. Okay. And then we assume that every voter votes for proposals that are closest to his ideal point. So I said one, you know, I want a 1.5 but I want to get Aaron so we vote two but you. So we vote on things two at a time, right? So you, you, you have the floor and you've done this and then we have to vote again. Then, then Aaron says. And two at a time means four against. Right. So then. So we're not ever voting for 1.5 or two, we're just voting for or against 1.5. No, 1.5 versus another proposal. Okay. Okay. So I propose 1.5 and you propose one because that's what you said you wanted. Right. And so I'm voting for 1.5 and you're going to vote for one. Right. And then, so then we end up with a two to one vote in favor of Trevor's proposal. Right. Because you wanted two and 1.5 is closer to two than one. His alternative is closer to my ideal point than your proposal. So he beats you two to one. Okay. Then we, then we can, then I, that's keep trying. Yeah. We can just run it over and over and over again. I put my proposal up versus yours. And then that, that beats yours two to one. And so we can keep voting and voting and it's easy to show in a graph, right, that the median voter's proposal will defeat all other proposals in a majority What does the median voter mean? In the conflict dimension, right, which is how much to spend on the fighter plane, the 50th percentile proposal, right, when we array proposals from a not much spending to a lot of spending in a left to right conflict dimension, the proposal in the middle is the, is the majority rule equilibrium. So that's me in the 1.5 too. And so it would eventually end up there, basically. So the hawk and the libertarian always lose, okay, the notion that a majority rule system, right, I mean the kind of cultural understanding that's come out of this median voter theorem is that, and you see it in the press, which is that basically some nondescript kind of preference in the middle that's, that the Cato Institute doesn't like nor does the Center for American Progress. That's actually where the political system inevitably has to end up under a majority rule system because of the median voter theorem. Now, is this, I mean, this doesn't seem surprising in the, in the sense that like, yeah, if you've got a whole bunch of proposals than the one that seems the closest to what most of the people would want is the one that's going to win out, doesn't have to be much, just literally the one in the middle that's in effect the least unacceptable, if you will, right? But is this then, I guess my question is, is this a bug or a feature of majoritarian democracy? That's the, that's the normative question. That's a normative question. He doesn't like normative questions. Well, no, it's just, I'm now in positive land and I'm not, if you want me to put my normative hat. But when, when people, so when this theorem was rediscovered in the 50s, you said, was this, was the outcome of it surprising to people? Did they, did it cut against what they had liked to believe or not? Because it just, I mean, what you've described just doesn't sound. You've described running to the middle, it sounds like. Based on what presidential candidates do, you got to run to the middle because you got to try and get to the middle voter, right? In effect, this is why. But I'm just thinking of like, if, if there was a theorem, if it turned out that the results of, so we have, all of us have our preferences and the results of a majoritarian system always led to the most extreme, but least widely held views winning out. That would seem to be a weird thing, right? Like that would sound like a bug as opposed to a feature potentially to a lot of people. I think one of the bugs here, there is a literature, I mean, in, in a public choice reader, there's a chapter on majority rule positive properties. That's what I'm going through now. And then there's a, the chapter right after that is majority rule and voting system normative properties. Well, one of these bugs could be, and this is, it brings in the libertarian point, I think, to some degree, but we'll get to that eventually, that you, let's imagine you had a very bimodal distribution, right? You had, you had people extremists on one side and extremists on another side and most of the people were there, which would be kind of weird, but it's, it's feasible, but like, and then still the median voter wins because it's the least unacceptable. But in a weird way, people, the best situation that would be for each of those people to break off and start their own country because it would, it would create, satisfy more of their preferences for them to just break away and let the group here and group here as opposed to both compromising in the middle, correct? There's a literature on that. The famous political theorist Robert Dahl, who recently died for it, who taught me and others at Yale, he spent a lot of time studying Belgium. Belgium, Belgium is for a democratic theorist. Belgium is an interesting case, right? Which is two, two linguistic groups, two cultural groups, two religious groups, they're trapped within the same border and they have to make collective choices together and it, this wasn't pretty, right? And so the, the, should Belgium break up? Should the South be separate from the, I mean, the, the, if you are trying to make collective choices with people whose preferences are very different from yours, you, there's lots of fussing. And there's a lot of disappointment involved. Oh yeah. Because nobody likes the result. It, it's, yes. So the marginal voter theorem is kind of a big initial thing in this, media is our, yeah, media and voter and big initial thing in this field. But we get to also Arrow's theorem and voting paradoxes and then it gets even kind of weirder, correct? Correct. So let me, so remember, so every, so the media and voter theorem works if all the conditions I just stated are true. So the one I'm going to alter right now is what if people do not have single peaked preference functions? And it's not just a mathematical curiosity. It, let me give a policy example. And it's easiest to do with defense policy. It was easy, two examples. One is the Vietnam War and one is the Afghanistan, actually. In Vietnam, there, there was discussion of, of some people wanted to new Kanoi and some people wanted to get out completely. And then others sort of the question is, what do we make of in defense spending and war considerations in particular? What do preference functions of voters look like regarding something in the middle? Remember, we, you just had a discussion of the median voter theorem kind of pushes things towards the middle. Well, what do we think of kind of getting involved in Vietnam or kind of increasing our involvement in Afghanistan, the so-called surge? Well, many voters have the following preference function, which is if you're a hawk, you really, I mean, the, the, the general Powell's, right? Theory of war is either you should get in with 400,000 troops and a gazillion tanks and, right, you should full force, full force, or you shouldn't get involved at all, right? So this is still in one conflict. So right. The conflict dimension is how much should we spend on military activities and country acts? And you could rank greatest spending the most, but your second choice might be none at all. And your least preferred alternative is a, is a middle level of spending. Well, then if you can imagine, I know it's radio and we're talking about an XY graph, but instead of an ideal point, and then all of the alternatives decaying inevitably in space as you get further away from your ideal point in the example I just described, you go down for a while, but then you go up again. And so that's like a camel that it's a two-hunk preference function. That kind of preference function yields to what's called a voter paradox, which is if we introduced spending proposals in the way we did earlier and we had someone with a non-syncopate preference function in, in the voting game, then X would defeat Y and Y would defeat Z and then Z would defeat X. And we have X beats Y and Y beats Z and Z beats X. That means majority rule would be intransitive. Go back to your ninth grade algebra class. Remember the intransitive properties of, or transitive properties of operations and arithmetic. That's supposed to be a property of rational people too. Like I like chicken more than beef and I like fish more than chicken. So I should like fish more than beef. Well here, individuals are rational. There's nothing irrational about anyone's preference, right? Some people don't like a middle level of defense spending for I think what are very defensible reasons. But the irrationality is at the collective level, right? Majority rule given non-syncopate preference functions cannot arrive at a result. So I think this is very strange. So we're going to have to work through this again in a, in a game style, I think. So with what I do in class is, I mean, I just show these kinds of preferences and then I show that X beats Y 2 to 1 and then Y beats Z 2 to 1. If the transitive property of mathematics applied, then X would beat Z. But in fact, when you do the voting, Z beats X. Can we describe that in practice with Trevor's fish, chicken and beef example by just assigning? Well, we have to have the preference function. So I think the idea is that Peter, Peter has this preference function, which, you know, A, better than B, better than C. And you are so, so Peter likes fish, better than chicken and chicken, better than beef. And you like chicken, better than, than fish. It works. It's most easily for our listeners to visualize that they put amounts of money. Literally how much do you want to spend on Vietnam or how much do you want to spend in Afghanistan and think about the possibilities as zero, a lot or something in the middle. And then imagine what your preference function looks like regarding those levels of spending. And I think some of our listeners would realize that their function would, would have a, a kink in it. It would, if you're a dove, you would like low spending the most. But you might not like a middle level of spending. You're saying, see, I understand the Powell rule. I mean, if we're going to get involved, let's go in. And so even if you're a dove, even if your ideal point is no spending, even if you're a Cato, a Cato follower, you still might have a preference function which says, gee, I'm in the minority. I mean, I'm not, I'm not going to win here. And so my second choice, more, more preferable to me than a middle level of spending or the surge is not a surge, but I want to go in with the first through 10th Calvary divisions and blow this country up. And one of the weird results of this that we never kind of get to an end because each thing keeps beating the thing before it is that if you're in a position to choose the order in which we vote on these things, you can, no matter which, what the public may prefer, you can pick which one's going to win by simply ordering the vote. What I just, what we just went through is what's called the majority rule cycle. And then Ero's theorem follows, which says, are there any solutions to this problem? Are there any collective choice mechanisms that solve this problem? And the, the, the result of Ero's theorem is, is, which by the way, was his PhD dissertation done in three months in 1951. And he won the Nobel Prize. And he won the Nobel Prize for, was that there are no, Everyone else can feel it out of you. Yeah, right, right. There are no collective choice rules that solve the majority rule cycle along with other sets of problems that are related to that. And so thus in the real world, we know that legislatures do make decisions. So how in fact, what's the relationship between the real world and these theorems that we're talking about? Erin's question answers itself, which is, once you realize what we've just learned, the order in which you vote on things then matters a whole lot. Because in real world legislatures, we do not keep voting on things to realize there's a cycle. We don't, in fact, legislatures understand what we're talking about. Senator Byrd from, from West Virginia was famous for the parliamentary rules and voting on amendments in certain orders. And the reason he did is because he knew some of the public choice theory without actually having been trained in it. He sort of understood that this is what was going on. And thus in the real world, since you don't bring up amendments that have lost, the order in which you vote on things can get you in the closer to your ideal point, if you're the majority leader of a legislative body, then if you vote on things in another order. And this comes up with a good example. I know you brought in about planes, right, the plane ownership example. There's a famous article in literature. Again, this is what I call the Caltech crowd sort of developed and worries about these kinds of things. And one of my, when I was teaching at Yale, the acting dean at the time was Mike Levine. Mike Levine was a USC law professor and hung out with the Caltech game theory crowd and then came east to be the dean of the Yale School of Management. And he also was an airplane enthusiast. And he was a member of an airplane club in Southern California. And he and Charlie Plot, who's a Caltech economist, social scientists, were both members of the club and they took all of this seriously and they arranged the order in which this airplane club they belong to considered they had to buy new planes. And so there was four possibilities and they brought up the budgets in various orders and they then wrote up an article in the Virginia Law Review describing their little mischief and basically, and I'll quote here from the article, since we thought that different good and good is in quotes here, agendas would produce different outcomes, we decided to adopt the good procedure, most likely to get the group to choose the fleet of planes we preferred, we being Mike Levine and Charlie Plot. And they did that by choosing the order in which they voted. And they rigged the agenda to get the status quo to be closer to their preferences than the preferences of other members of the group. And anyway, this is legend in the literature is sort of proving convinced and then they did laboratory experiments to show the same thing, which is the order in which you vote on things in a real world legislature and determines the outcome that you end up with and thus majority rule in the real world is not neutral in that sense. Were they asked to leave the club after this? The article does not reveal. When I've heard this story before, I'm always struck, I don't know by its implausibility or I mean because it sounds like this thing like they pick the order and then suddenly the planes that they wanted, I mean it ends up sounding like in heist movies, you spend the first three quarters of the movie watching these people do all these things that you don't quite understand, setting up this heist and then all of a sudden the last bit of the movie is like all of it falling into place and this like brilliant sequence of events and they emerge victorious and that's what this sounds like. This sounds like a heist movie. I think they wanted it to sound like that. But this is even still this is a very small case in the sense that there's a small number of people in the club. There's a small number of options. It's easy to figure out which ones they want. Is it realistic to pull this sort of thing off at say a legislative level where you've got many, many more people that it's just it's going to be harder to figure out exactly what everyone wants, arranging the number of sequences you'd have to go through would be much higher. The point of this is not that it's not that remember arrows theorem, right? The bottom line was the only social choice aggregation rule that salves arrows paradox is is dictatorship. So what the subsequent literature in this genre has tried to do is tried to sort out how vulnerable majority rule was to agenda control. That's all it. So it's not that speaker Boehner can get what he wants. But the committees and the order in which things are brought up and then things are sent to committee and they don't come out of committee and all of all of those things imply that the kind of high school textbook notion of majority rule as fair to all preferences under all circumstances. And thus we all should accept the result because a fusion of normative and positive thinking what this part of public choice literature in effect does is make one realize that the ability of majority rule to come up with decisions is rather fragile and that under under a reasonable set of circumstances the results either cannot occur and therefore the extent they do occur that means they reflect some people's preferences more than other people's preferences and thus the aggregation procedure is not neutral and that that that's the point of of this literature and so since a since so much of libertarianism is about the the notion that the public sector should be as neutral as possible with respect to people with respect to preferences and goals and all that because we're also different and libertarianism more than other political traditions understands how different people are and thus wants to minimize the use of collective choice in general one can understand why one would become I mean the more you study the properties of collective choice procedures the more you get scared by how they can be manipulated by those in charge so ironically even though the caltech stuff is very mathematical and very abstract and very dry and hard to picture and it has no obvious emotion connected to it I think it fits quite easily into libertarian thinking because of what I just what I just said earlier on you had distinguished the the Virginia school of public choice from the caltech school of public choice and I think everything we've discussed so far has been largely from has been caltech sorts of things so I was wondering if you could now what what are some of the contributions of public choice from the Virginia school well the most common term that comes out of it that Kato folks often use we use in our all our writing and and and podcasts and things is rent seeking where the notion that there are market privileges out there or that what what the legislature does is in part is create exemptions for various folks from market results and that one then is willing to pay for that exemption and thus the campaign contributions and the flows of money that occur we now use the term rent seeking to describe and then then there's a literature on not how much money is in politics from from this point of view but actually how how little I mean if you if you realize how valuable entry restrictions or subsidies really are to various interest groups what stunning is not how much money there is in politics but how little because the the the economic privileges that the that the legislature gives out are quite valuable now it seems that this is tied although we we've discussed that they're just new schools especially this idea of what you can get out of politics and how you get things out of politics that the characteristics you talked about in terms of voting they seem to mostly emerge from the lack of intensity lack of the ability to to vote harder on certain things which which is many people think is the biggest virtue of voting one man one vote type of democracy is is the great leveler of human preferences because one man one vote but of course if you just constrain the choice mechanism to one vote it you haven't actually eliminated the intensity preferences correct and so then you can get weird things so this all seems to go into rent seeking and political spending and all these other things that go log rolling trading because ever everyone is not now binary like there's people who don't really want something and do really want something else and they have one vote to give and so then they start trading them. Okay so go back to your kind of microeconomics and remember the term Pareto optimal right that that under normal circumstances that we teach about in class markets result in gains to trade for everyone and then don't redistribute so one once markets are at equilibrium all gains to trade have been exhausted and there are no other possibilities that exist and that's called a Pareto optimum. A puzzle and for economists is why do legislatures pass so many programs that actually aren't Pareto improvements in other words in economics we've discussed in a previous free thoughts discussions that there are things called public goods there are things that markets aren't that good at producing one can conceive of collective action as a way of improving upon what markets do in class or in black on the blackboard but in the real world. Legislatures actually are very bad at producing real public goods and they're really good at producing what are you might call locally privatizable stuff that governments have to do and so the ultimate of this this insight or this thinking there's a there's a paper in the literature by Emerson new and Peter Orteshuk 1985 and it's it's it's a puzzle about why so many bad projects so many bad things are are passed by legislatures that don't actually improve national welfare I provide public goods that really can't be provided by markets and all that so this another way of describing this is why does Congress produce so many pork barrel projects that are that have very district specific or state specific gains and then the charges are to the rest of the country that don't get any benefit from these local projects at all and new and Orteshuk works through a game in a mathematical model and they come to the conclusion that the problem is that we have geographic districts remember we so we so public choices about how can we aggregate people's preferences into collective choices that makes sense so one could conceive of a national legislature as having a national vote right there are no states there are no localities there aren't there's just one big club 330 million and we vote for people and then there's something called the legislature and they decide on things well kind of scratch your head and realize we probably wouldn't have any dams in in Alabama funded by the Army Corps of Engineers if we had that kind of collective choice procedure new and Orteshuk basically put the mathematics on this and then say believe it or not a big source of mischief in the United States political system which and I'm not attacking federalism here but ironically having geographic districts used to create a national legislature leads to vote trading where they conclude the result is something called universalism which is simply stated I'll vote for your stuff if you vote for my stuff i.e. look things with very geographically defined benefits and diffuse public costs and every district and state has a list of all these things and then when they go to Washington everyone realizes we could vote against all these things which actually would be Pareto optimal because none of them are true national public goods or we can vote for all the wish list of all of the stuff for all of the stuff it seems like a prisoner's dilemma we could even everyone has to vote against it Aaron is no good student shifts to Trevor you're exactly right it is a prisoner's dilemma game which is if I observe that I vote all these things down for my people but you get your stuff that's not an equilibrium and then everybody has to get into it right even if you're libertarian I mean my great example is mass transit spending all the spending initially in the seventies went to New York and Chicago we didn't have and Boston I mean now my little my little town where I grew up in Northern New York Watertown New York has a bus system has 20,000 people and it has a federally subsidized bus system there's now mass transit everywhere because I mean it's not an equilibrium to have a federal program where three cities get all the results right so it's a great example of embers of new and of universalism which is if and we have it because we have geographic districts in a national legislature but we also have it because the preferences are very underneath the vote and so you said that so you are willing to give away but the appropriate right the these I mean again the despite the claims of members of Congress the true beneficiaries of the big dig in Boston right that huge thing to put the 95 and 93 underground that Tip O'Neill got through Congress in the late eighties which took a gazillion years to build a gazillion dollars and a gazillion right the benefits go to Boston I don't really feel them here in Washington where or Lord or people in you right but every the whole U.S. is paying for it well then the Utah people say gee I need we need some better access to to the ski lifts and so Senator hatch make sure that some of that stuff gets made so it's so everyone has to cooperate and say no and you were quite right to describe it as a prisoner's dilemma and that's not an equilibrium eventually someone defects and says this would be really good for my state and the rest is history and why don't the voters get I mean because they're they're the ones being pillaged by all these they're pillaged a little right so that so this is that that term that we sometimes talk about concentrated benefits and diffuse costs and so the voters getting that military base put in with lots of jobs is you can see these huge benefits from that but the handful of pennies or dollars out of your tax bill that went to putting in a base or a dam or a bridge somewhere else you don't really notice. I mean give Amtrak as an example right make believe Peter I run for office and I make the following pitch to voters right so Amtrak is a billion something a year that's basically $10 a tax paying unit okay per year in the United States so make believe I run for office not my commercial say if you elect me I'll give you back $10 a year okay and then think of the Amtrak employees they make 80 to $100,000 a year they get annoyed speaking of intensities right they start they give to Pax and under Cato approved political expression they give a lot to the new groups that can be that can say and they'll have ads on TV to talk about there be no trains. When it's not just the employees there are also there are people who take Amtrak and benefit greatly from the reduced costs the buying the fact yeah so most people in the country do not take Amtrak but the $10 doesn't matter much to them but if you're someone who commutes every day on it even if it's not you're not getting your paycheck from them but they could do that the real losers of the employees but the employees won't talk about themselves in the ads will they they'll talk about the riders they'll try to gin up the riders right the riders have car they could do other things right it's not mean but the real crisis is for the the people who work for the firm take this a program at a time and you can see why spending cuts always lose. So the only way to solve this is the public goods part right it's a collective action problem where we all have to join hands and say I'll give up my stuff if you give up your stuff right and that's called the political party next actually called libertarianism I mean ironically what small government people are trying to do is solve a collective action problem where we all agree nobody gets anything. What if you introduced the spending cuts not piecemeal like we're going to get rid of Amtrak and we're going to get rid of sugar subsidies or whatever but you said we're going to cut all of these here we put up for vote cutting all of these programs that would add up to thousands of dollars for each taxpayer correct and we've talked about this before the base closure and realignment act that that's the one time Congress recognize some of the game theory that we're describing today is defense spending they know that individually base closures in their districts are death they know that and so now and then they've come up with a prisoner's dilemma correction procedure which makes the nation better off even though it hurts each particular area where a base is closed and we could you could think of generalizing that concept which is what you just did to all sorts of geographically concentrated spending of the sort we've been talking about under those circumstances you could conceivably get you'd have to have the key is no amendments the key is also that the set of geographic areas that lose are a minority of the legislature so if you're scaling it up right your proposal if everyone's losing if a majority of districts realize their losers then I mean because of the progressive income tax and because so much of our public spending actually comes from four states the tax the taxes to pay for things come from Massachusetts New York New Jersey and California you'd be stuck right if you haven't got the percentage of that I'd have to do some dig work but it would it would stun you at how big a percentage of national tax revenue comes from those areas it goes to everybody else in the country so to implement Aaron's report that if we try to do this for all programs at the same time and I'm not talking about Medicare you know the kind of other but things that have geographic public goods characteristics the rest of country would realize that this isn't again so even in a true prisoners dilemma based closing kind of situation Aaron's proposal lacks a crucial ingredient which is the set of geographic losers have to be a minority of the legislature and once everyone realized that majority of the area of the country benefits from all of this stuff because the money comes from so few places then even a prisoner's dilemma solution would not cut spending if we tried to implement it so there seems to be something oddly mechanical and almost inevitable about this if you say these are the systems these are the kind of things they reward this is the way you can disperse costs this is the way you can voice costs upon other people within this voting system you kind of seems to predict the whole the whole thing going forward so this vote these voter theorems that we've discussed today and public choice stuff it doesn't just lead to the conclusion that voting is not very a very good way of aggregating preferences and we should prefer other ones is that the moral of this whole story is that you vote about too many things you have too big of government giving way too many things with too much power and eventually you're going to get a prisoner's dilemma insoluble game where everyone's just taking everything from everyone constantly and of course we hear constantly all over the country that Washington is broken because they're kind of saying everyone's just giving every way to their friends and that's what we have because of these theorems so is that is that what you're willing to say for this voting bad markets good and what makes that if that's true what makes that the case the problems even though today's discussion has been about all the problems with collective choice the alternative of well there's two alternatives won't generic alternatives one is no government the second is different kinds of preference aggregation systems that allow intensities to be part of the system and there's board account there's single transferable vote there's all there are there's a whole literature on schemes that try to allow intensities to play a bigger role so that those that were highly taxed in effect could put more in the game so that they weren't so highly taxed and say I don't like this but then on the opposite side there's there people who gain from the redistribution they gain a lot from the redistribution and they have intensities as well they just don't have wealth so we're back to the normative discussion which is what is the purpose of politics and public choice textbooks the good ones talk about well that's actually what we're struggling over which is one more mechanistic view kind of from the heart of economics is that markets do most things well they do certain things not very well the state kind of could in theory do those things well public goods things like that in practice we've just described why they're not very good at that at all and so now what do we do the answer and the end of my lectures I usually say I don't know because either you get rid of the public sector or you try to constrain it with constitutions which we've tried to do and we have a Kato so much about how that should be an important part of the public sector but in practice having nine people dressed in robes try to say no to majority rule systems is not it's they don't have an army and so all they can do is kind of squeeze it and nudge it a little bit but if they try to say no to everything that legislatures want to do that wouldn't be an equilibrium we find that out in 1938 right that that in effect we could say we wish 1938 didn't happen but I think if it would happen inevitably all over again and it's this tension between many people outside the Kato building think that what the public sector does which is redistribute is actually what the public sector should do but then you have the other problem that they also think that intensity of preferences allowing that to have the effect especially when money is involved is wrong is wrong yeah which is against what so in this in this building will often be like well no like that if you want it you should pay more for it but they're increasingly we're getting the idea that that's that's a wrong way of looking at it and voting is better but of course voting is not exactly egalitarian in that way is it kind of what we talked about today when this comes up with campaign finance right which is the for the one person one vote folks the enormous expenditures by the affluent trying to get the rest of us to listen to them more about their preferences is normatively wrong from their point of view and from our point of view we just see it as an extension of free speech so I think actually that that's I find campaign finance to be the arena where the normative struggle over what democracy is about it actually is being played out and probably will go on forever and ever never because once you choose which way I mean I think having rich people try to tell us what they want and why I don't I'm not worried about it because in the end everyone can still vote the way they want and often the rich are very frustrated by no one listening to them they do not get what they want even though the ads that they pay for are so strident about their views I think what's interesting to me is so I think that's I think what the Supreme Court has allowed which is to have money-backed free speech is actually a good antidote to some of the things I describe but in the end people are still free to vote the way they want and I probably am a Democrat with a small d at heart and so I really don't want to constrain the ability of people to vote and have their preferences even if I profoundly disagree and think that they're wrong and misguided etc etc I worry very much about in effect the dictatorship the kind of arrow talked about as the only solution to some of the dilemmas that we have that that's bad too thank you for listening to free thoughts if you have any questions or comments about today's show you can find us on Twitter at free thoughts pod that's free thoughts pod free thoughts is a project of libertarianism org and the Cato Institute is produced by Evan Banks to learn more about libertarianism visit us on the web at www.libertarianism.org