JBvPsj8H

1A THESIS ON THE ROLE OF GOVERNMENT
2James de Vries
3
4INTRODUCTION
5The role of government is a highly contentious issue. Government is constantly pulled between the competing values of its constituents - values which often come into direct opposition. This is a problem I fear human civilisation will struggle with indefinitely, wherever humans with different values live under the same code of law. For this reason, it might, and indeed it very much should, seem extremely presumptuous for me to claim knowledge of the "proper" role of government. However, this problem is greatly mitigated by the capability of economics to strip subjectivity from the equation. Wherever two humans desire two conflicting things in a market, the matters of how much they are willing to sacrifice to fulfill that desire, and how difficult those desires are to grant, are elegantly weighed against eachother.
6
7Before I proceed any further, I wish to take a moment to address the reader. A certain number of you will curl your faces up in distaste at the mention of markets, and the use of markets as the primary vehicle for maximising personal utility. This is understandable, because while advocates of free markets make many truthful claims about their power and efficiency, they are also known for spouting the most ludicrous nonsense, to the point where some think that the state can be done away with altogether. So, while I understand the skepticism I may have generated inside you by taking this approach, I implore you to see this thesis through to its conclusion - and offer you this promise: Even though we may disagree much of the time, there is something in this thesis to enrich the value proposition of almost every political ideology.
8
9Communists will find themselves at odds with much of what I have to say, as much of what I have to say revolves around the administration of markets. But even communist states have systems of crime and punishment, and this thesis offers a powerful lens with which to look at the problem of crime and punishment, in a way that ensures minimum injustice. They may also find themselves agreeing with many of my critiques of modern capitalism, while disagreeing with my responses to those critiques.
10Social democrats may raise skeptical eyebrows at how much emphasis I place on productivity, markets, and economics, but will ultimately be pleased to find that this thesis offers them a way to justify many of their most compelling proposals from first principles, without having to appeal to humanitarianism. They will find this helpful, since many ears are deaf to humanitarian concerns.
11Libertarians will, provided they do not subscribe to a deontological morality which considers freedom from taxation as more important than even their own personal wellbeing, will also find much to like. The substance of the thesis provides a powerful bulwark against authoritarianism, as it proposes a tightly bounded and clearly defined sphere to confine government within - though that sphere may be bigger than some may like. Being able to agree on exactly what role government should be restricted to may be useful in uniting various libertarian factions.
12Many fiscal conservatives will find it controversial when I claim that sometimes, investing in other citizens - what might be termed "welfare" - is actually in the best interests of the very taxpayer from which the funds for that welfare are taken. But I hope that their rational self-interest will trump whatever affront this might be to their existing worldview.
13
14There are only two classes of political ideology which i'm aware of, that will find nothing of value in this thesis.
15
16The first, is anarcho-capitalism. This is because anarcho-capitalism is incompatible with the economic reality of oligopoly, and by extension incompatible with economics as a whole. Anarcho-capitalism suggests that the states monopoly on violence should be broken, and replaced with a diverse portfolio of violence providers, but offers no way to prevent this diverse portfolio of violence providers from forming an oligopoly. If they choose to form an oligopoly - and indeed they most certainly would, given that there is no greater incentive to do so than having total control of the entire economy and everybody in it - then the state would be reborn, and anarcho-capitalism would end.
17
18The second class of political ideology which will find nothing of value in this thesis, is authoritarianism. This is simply because authoritarian ideology is predicated on the idea that the state knows what is best for people, better than those people themselves do. Not only is this intrinsically dangerous, and at odds with the concept of consent, but I am yet to see any evidence that it can be true without unpersoning members of your society with heterogeneous values - something that I consider to be undesirable. The moral anti-realist in me wants to apologise for leaning libertarian, since it's "just the way I am" - but the fact of the matter is that respecting personal autonomy is simply the best way to reconcile the differences between human beings. I might go so far as to say liberty is instrumental in all plausible strategies for maximising utility in a human population.
19
20In this thesis I examine the current implementation of capitalism as well as the functions currently performed by governments in those economies, and assess the legitimacy and viability of this current state of affairs by looking at governments as tools for facilitating markets in solving the problem of utility maximisation for their populations. I also propose alternative solutions to the problems these functions address, where the solutions presented by government are evidently sub-optimal with regards to utility maximisation.
21
22I hope now that, whatever your political persuasion, you are prepared to give me enough time to justify my claims. But even if not, there is little more stalling or hedging to be done.
23
24THESIS
25It is as follows: 
26"The role of a nation's government should be only to maximise the efficiency of its economy"
27With the corollary: "this applies even to domains where no market exists or can exist".
28
29Madness! What about rape and murder? What about education and healthcare?
30
31These are all good and righteous questions to be asking. They all have more or less the same answer, and to explain it in complete detail will take the entire length of this thesis. The short answer to these questions is that rape, murder, disease, and illiteracy are all inefficient and costly, and that "restoring efficiency" to these things looks very similar to the justice system, universal education, etc that we're all familiar with, barring some important differences. Despite the fact that this thesis will measure utility in dollars, that does not imply that things not traded in markets are percieved as worthless. On the contrary, the above thesis should be the most boring and least contentious statement ever, since it essentially boils down into "make the most good stuff happen using the least scarce things" once you substitute in a reductionist definition of "efficiency" - that is, generating maximum utility from the available scarce resources. Putting it like that makes it sound truly vacuous, but I assure you, despite how tautological the thesis itself may sound, viewing the problem of governing through this purely economic lens allows us to make some damning diagnoses of heretofore mysterious social ills. And it also illuminates some seemingly unintuitive paths forward for our society.
32
33But before we will be able to get into any of that, I must first detail a little about what markets and efficiency ARE. Efficiency is essentially the property of generating as much as possible of some desirable thing, for as little cost as possible. Good and bad, desirable and undesirable, are wholly subjective. And yet we are trying to maximise "desirable things" across the entirety of a heterogeneous population. The term for how useful or desirable a given individual finds something is called "utility", and every person obviously interprets different things to have differing levels of utility.
34
35I alluded to the solution to this earlier, and indeed the solution is very old news - the solution is markets, and markets have existed since pretty much the start of recorded history. I don't expect many people to require a justification for the necessity of markets, but I would like to be as inclusive as possible, so I will still briefly outline the case for the sake of any readers who are for the abolition of markets or do not understand their importance. One of the things this thesis accepts as axiomatic is that the individual is generally the best authority on what is "best" for themselves, and that the individual will act in such a way as to get what is best for themselves. This isn't always true, but it's much truer than the alternatives. Because every individual has a differing "utility function" or set of desires, its necessary for each individual to have some say in how scarce resources are allocated, so that they get some utility from those resources. However, one-person-one-vote systems are very problematic, in that they do not differentiate between different magnitudes of utility across individuals.
36
37Say there is a voting body, comprised of nine people. Let's also say that 4/9 members of that voting body want a resolution to be passed, and they will it to be passed so powerfully that they would gladly sacrifice their lives for that cause. Then let's also say that the other 5/9 members of the voting body lean slightly against passing that resolution, and they only have a relatively weak preference for it to be struck down - perhaps comparable to how they might prefer to have a steak for lunch today instead of a salad. It should be self-evident that in this case, the majority needs to step aside and allow the four to have their way, simply because the passing of the resolution doesn't really seem to concern or impact the five particularly much compared to the four. But when they lean against the resolution, and vote in their own self-interest (as every system should expect them to), they bring the entire power of their 1/9th representation to bear on the issue. Differentiating between "disagree" and "strongly disagree" is also futile in this instance, since they have no incentive to be honest, and still have the option to bring the full power of their vote to bear against their opponents at no cost to themselves. The only solution that can reliably extract an honest assessment of utility from a human being, is to determine how much they are willing to sacrifice in exchange for that utility. And if you have a system of voting where the power of your vote is variable based on how much you are willing to sacrifice, you need only for those sacrifices to be of utility to the other participants, and you will discover you have re-invented the market.
38
39And that is ultimately what market economies are - a decentralised network of people and organisations voting on the allocation of scarce resources, where the sacrifice of something valuable solves both the problem of all votes being equal when utility differs, and the problem of authentication. The way it solves the problem of all votes being equal when utility differs should require no further explanation - I'm sure all readers live in countries with market economies and are familiar with the concept of an auction or bidding. And the way it solves authentication is that when your vote is not authorised by your identity but by some sacrifice or contribution you make, authentication is unimportant. So long as they make the appropriate contribution of labor, resources, or a currency that denominates those things, people are quite welcome to cast the same vote as many times as they like, anonymously.
40
41This is not a new way of looking at markets. But I remind you exactly what markets are, because it's important to remind you of some things that markets are not. The particular brand of capitalist market economy that is common today has a lot of strange idiosyncrasies which nobody seems to consider strange anymore; they have become part of the furniture so to speak. And this is made worse by peoples habit of anthropomorphising this idiosyncratic, man-made system as though it were something fundamental, a universal constant.
42
43These idiosyncrasies include, in no particular order:
441. That land rental yields substantially exceed taxes on land values.
452. That certain actions are almost always not allowed (such as stealing) and other things are almost always allowed (such as contaminating the shared atmosphere).
463. That certain speech is almost always not allowed (such as libel, and death threats) and other speech is almost always allowed (such as talking about the weather, or your political opinions).
474. That certain things are almost always publicly funded (such as police, fire and rescue), while other things are more rarely publicly funded (such as tertiary education and insurance), and other things are almost never publicly funded (such as software).
485. That information property has a specific set of protections (copyright law, trademarks, patents, etc)
49
50I will talk in turn about each of these idiosyncrasies, what the situation currently is, what contention there is in the discussions surrounding this issue (if any), and what the solution to the problem would be if the government were playing it's proper role, by maximising the efficiency of its economy. You will find that in all mentioned cases of arbitrary lines between admissible and inadmissible actions/speech/spending, the arbitrary line is usually a flawed approximation of efficiency, and that there is value to be found in re-drawing that line accurately and clearly based upon this principle.
51
52Just a few points, before we get into the meat of these issues.
53In this thesis, I will be measuring utility in dollars. That is to say, desirability, happiness, whatever you call it, it is called utility here. Utility will be measured in dollars, which represents the dollar value the individual considers equivalent to a product that grants that amount of utility. The astute reader will have already deduced the problem with this approach: This way of measuring utility will be distorted by any inequality in purchasing power. This will mean that we overestimate the utility the wealthy get from their purchases, and underestimate the utility the poor get from theirs.
54
55This is fine for my purposes, for a few reasons:
561. In the real world, data on underlying utility is extremely scarce. Sales data, on the other hand, is overwhelmingly plentiful.
572. Even though it might be possible to describe the curve of diminishing returns where greater wealth increases happiness less at high levels of wealth, constantly transforming back and forth between "dollars of utility" and "real utility" would make the worked examples in this thesis tiresome to write and difficult to read. Suffice to say that, since we intentionally err in favor of the wealthy, social policy prescriptions in this thesis are better thought of as a minimum, or "floor", rather than a direct recommendation.
583. I think it will make it easier for me to earn the trust of readers who are skeptical of social welfare. Rather than performing questionable, presumptuous, and barely-scientific number crunching to try and normalise utility across wealth classes, I think skeptical readers will find the thesis more credible if I am able to justify my social policy prescriptions "with one hand tied behind my back" in this way - and I will do so.
59
60Let's begin analysing some of these idiosyncrasies, shall we?
61
62ON PATENTS
63It is hard to think of another idea as half-baked as the modern world's conception of patents. There obviously needs to be some kind of reward for being the one to invest in researching a new technology, or else R&D funding would dry up. Also obviously, that reward has to scale based on how well received that technology is by the market, because there should be no bounty for useless gimmicks. But why on earth should the reward be state-enforced monopoly? Monopolies are monstrously inefficient. Every developed country has competition or antitrust laws specifically designed to prevent mono- and oligopolies from forming, and destroy them if they do, for that exact reason. And then alongside these laws, they also all have entire government departments dedicated to creating and enforcing artificial monopolies, with the power of the state behind them! It's madness, really, when you think about it. It's an idea that was clearly come up with on short notice some 250 years ago, and we're still just running with it today.
64
65I will take a moment to explain the exact problem with monopolies, for those not familiar with the principle. The following is a simplified but still lengthy explanation of the conflict between profit maximisation versus utility maximisation. If you already fully understand demand elasticity, and how monopoly affects pricing, feel free to skip ahead until you see a paragraph which doesn't have any numbers in it.
66
67Imagine that there is a technology under monopoly, which cost $5,000 to develop, and a population of consumers who have a varying level of desire for this technology. For simplicity's sake lets say we have a population of 99 consumers, and the population is divided evenly across three categories:
681/3 of the population do not want the technology particularly much. We will call these "Reluctant consumers". They think they would derive $100 of utility from it, and so will not buy it unless the price is $100 or less.
691/3 of the population find the technology intermediately desirable. We will call these "Regular consumers". They think they would derive $200 of utility from it, and so will not buy it unless the price is $200 or less.
70The final 1/3 of the population find the technology extremely desirable. We will call these, "Eager consumers". They think they would derive $400 of utility from it, and will only buy it if the price is $400 or less
71
72The cost of producing the product that contains this technology is subject to economies of scale, like all mass production. Lets say that the product costs $50 per unit to produce, when selling to 1/3 of the population, $40 per unit when selling to 2/3, and $30 per unit when selling to the entire population of 99.
73
74Lets look at how much utility, and how much profit, is generated at these three price points.
75
76When selling the product for $400 per unit, we make 33 sales, and our costs are $50 per unit. This results in revenues of $13,200, costs of $1,650, and profit of $11,550. This is the most profitable price point. The amount of utility generated for the consumers, approaches zero. The eager consumers were the set who valued the product at exactly $400 of utility, meaning that for them, the existence of this product did not allow them to gain any more utility from the expenditure of that $400 than if they had spent it elsewhere. They essentially gain $400 of utility from the product, and lose $400 of utility from the money they spent on it.
77So, the consumer base hasn't gained any utility from this technology. But society has; these products generated $13,200 of utility for $1,650 of production costs and $5000 of research and development costs. This represents a saving of $6,550. On the whole, a lot of scarce resources were saved, for generation of the same amount of utility than if the technology was not developed. When scarce resources are saved, everybody in the society benefits.
78
79Let's consider the next price point. When selling the product for $200 per unit, we make 66 sales, and our costs are $40 per unit. This results in revenues of $13,200, costs of $2,640, and profit of $10,560. This is slightly less profitable than the last price point. The amount of utility generated for the consumers is as follows: 
80The regular consumers gain no utility. They gain $200 of utility from the product, and lose $200 of utility from the money they spent on it.
81The eager consumers gain $200 of utility each. This is because they gained $400 of utility from the product, and lost $200 from the money they spent on it. This results in a net increase of $6,600 across all consumers.
82But much more importantly, the social benefit: These products generated ($200x33 + $400x33) $19,800 of utility for production costs of $2,640 and development costs of $5000. This represents a saving of $12,160 of scarce resources to generate the same amount of utility.
83
84Then the final price point. When selling the product for $100 per unit, we make 99 sales, and our costs are $30 per unit. This results in revenues of $9,900, costs of $2,970, and profits of $6,930. This is by far the least profitable price point.
85The reluctant consumers gain no utility, as usual.
86The regular consumers gain $100 of utility each ($200 gained - $100 spent) for a total of $3,300
87The eager consumers gain $300 of utility each ($400 gained - $100 spent) for a total of $9,900
88This adds up to a net increase of $13,200 across all consumers.
89
90But here's the kicker: These products generated ($400x33 + $200x33 + $100x33) $23,100 of utility for production costs of $2,970 and development costs of $5000. This represents a saving of $15,130 of scarce resources to generate the same amount of utility.
91
92And that's why monopolies are terrible. Because monopolies are inefficient. Without competition to drive the price of the product down to near the cost of production, the price point that the monopoly will choose is always the one that generates the highest amount of profit, regardless of how much inefficiency that generates. With competition, the lower price makes the technology accessible to more of the people who can put it to good use, and also often means that more economies of scale can be leveraged in the production process. If you want to "make the most good stuff happen, out of the least scarce stuff", competition is not optional. Well, not in capitalism anyway.
93
94None of what I just said is news. It's ancient, common knowledge. The first antitrust laws were passed in 1890. It is at that point, not today, that the mechanism of using temporary monopoly as a bounty for R&D should have been re-examined. The stark difference outlined above, between the efficiency of a lower, competition-driven price point, and a higher monopoly-driven price point, grows even greater when demand is inelastic. Demand inelasticity is a term that essentially means, people do not buy more or less of something based on its price. If you have a life-threatening illness, and you need a drug for treatment, you will buy exactly as much as you need, regardless of what changes there might be to the price of that drug. In situations like that, the inefficiency of monopoly is amplified, as the price people will pay ceases to be bounded by percieved utility and is instead bounded only by the wealth of the consumer. It is in situations like these that inefficiency ceases to just be infuriating to economists, and instead spills out to garners the fury of the wider population. It's hard for regular people to get mad about bad economic management when it's just a numeric footnote to a seemingly-functional system. It's easy for regular people to get mad when diabetics die from monopolistic price-fixing. But they should be just as mad about both, because they are both artefacts of the same phenomena - inefficiency. Sometimes inefficiency is one person dying. Sometimes inefficiency is a million people spending an hour in traffic. But both of those add up to a lost lifetime. You will never know how prosperous we could have been, were our economy efficient, just like you will never know how many life-enriching experiences your time spent in traffic has cost you. The only way to find out, is to fix the problem and see.
95
96Speaking of fixes, I present a very simple one. Not because I think its a perfect solution, or even a good one, but because that's how little thinking I had to do to improve on the current system. Consider this: Instead of patents granting exclusive rights to the commercialisation of a technology, instead, all products which use patented technology are subject to a levy on all sales revenue, paid to the patent holder. This would be no more challenging to enforce, because all people using patented technology needed to be identified under the old system anyway - just for prosecution instead of levying. This levy would last for a comparable period of time as the current monopoly periods do. Discussions on what duration/size of levy would be optimally effective, is beyond the scope of this thesis. But even though the ideal levy might be extremely difficult to determine, you could get a "good enough" levy simply by wiggling the knobs up and down until R&D spending looks similar to the levels they were before the change was implemented.
97
98If the government was properly performing its job of maximising the efficiency of its economy, then the holders of patents would still be subject to competition, and would be compensated in some other way. The damages of this current implementation of patents are sometimes even counted in lives, but more frequently it can be counted in the dollars which are not earnt from the provision of utility, but rather from the exploitation of state-enforced monopoly.
99
100ON OTHER INTELLECTUAL PROPERTY
101Much like patents, much intellectual property constitutes a micro-monopoly. Again, much like patents, a bounty of this sort is necessary for the monetisation of the intellectual property. As the development of intellectual property provides utility, it's important that it's development is sufficiently profitable, otherwise less or less valuable intellectual property will be generated. This is unavoidable; but the problems with monopolies do not disappear just because they are better than the alternative. They are still just as inefficient. Intellectual property has no overhead costs on replication, and it provides utility. Theoretically, this means that if there was some way to remove all restrictions on intellectual property without destroying the incentive structure for its creation, it would be infinitely efficient to do so.
102
103To this end, I propose a system that would allow us to exploit this "infinite efficiency" of free replication of intellectual property.
104
105As it stands, most intellectual property follows an exclusionary model like what we have with patents. Somebody creates some intellectual property - be it a piece of software, or an original artwork - then the circulation of that intellectual property is restricted as much as possible, to create artificial scarcity. This is a necessary evil. Unless the intellectual property is scarce, we cannot use the market to determine how much utility that intellectual property offers. Without a market determination of utility, there is no way to determine the appropriate reward for the creator. And unless the creator is appropriately rewarded, intellectual property will be under-produced, creating inefficiency.
106
107However, it may be possible to get some of the best of both worlds - the world of markets and scarcity, and the world of free information and infinite efficiency.
108
109But before we can get into that, I must furnish the reader wikth some background on externalities. The ability that is unique to government - as opposed to other economic entities like individuals and corporations - is the ability to appropriately invest in things which generate large, difficult-to-capture, positive externalities. As a rule, these things will be underinvested-in by private enterprise, by the proportion of the externality which cannot be captured. So if say, half of the benefits of privately owned freeways escape capture by its builder, then there will be about half as much freeway investment as would maximise utility given the available scarce resources. An important addendum to this fact is that the process of capturing the externalities itself reduces the efficiency of the investment by both making it more costly, and reducing utilisation.
110
111That's a bit of an esoteric statement to just throw out there by itself, so let me illustrate with some examples. If we continue with the freeway example, that freeway might be utilised by 10,000 vehicles a day, or somesuch, under the circumstance where it is free. This is the highest utilisation the freeway can achieve, because demand is inversely related to price, and when it is free to use, price is zero. Now if you build toll booths on the freeway, this is already inefficient before you turn them on - because the toll booths do not provide utility, but have nonzero cost, building toll booths increases the cost to utility ratio of the freeway and reduce its efficiency. But more than that, because of demand elasticity, when you turn them on, the assets utilisation decreases. People decide they would rather take a less-efficient-but-free route to their destination, than an efficient but costly one. This causes the utility provided by the freeway to drop even further, because now not only is it more expensive, but it is serving less people. More scarce resources are consumed, and less utility is provided overall. Your attempt to capture the positive externality of the freeway's use has created an inefficiency double-whammy.
112
113Now, this is where you might ask me: "If freeways are so much more efficient publicly owned versus privately owned, why stop there? Why not just nationalise all industries and resort to socialism?"
114
115It's a good question with a good answer. And the answer is, "Because for an investment to be made more efficiently by the public sector than the private sector, an outsized proportion of it's benefits must be external". Let's say you decide to make a public investment in something arbitrary like, say, a farm. You have no reason to expect it will provide more utility than the cost of its construction; because all of the benefits of a farm are internalised, market forces will have brought the cost of its construction roughly in line with its overall utility. The class of goods and services which are most efficiently provided by government is exclusively those goods and services with high net positive externalities. This includes things like:
116
117Fire departments, because your neighbors fire being put out protects your own house.
118Police departments, because your neighbors thief being arrested protects your own property.
119Garbage disposal, because your neighbors garbage festering in their front yard will put your own household at risk of a plethora of filth-related dangers from bacteria, moulds and pest borne diseases.
120As well as any good or service where the efficiencies enabled by it's public nature outweigh the inefficiencies of inaccuracy in the estimation of utility the good or service will provide. This includes things such as non-discretionary health insurance, infrastructure, and - I will argue - it SHOULD include intellectual property.
121
122We have already established how the proliferation of intellectual property is "infinitely efficient" due to its nonexistent cost per unit, and nonzero utility. But that is not enough to warrant the removal of intellectual property protections, because it is not enough for there to be efficiencies to be gained to make it a good idea; there is a second predicate. It is only a good idea if the efficiencies that stand to be gained exceed the inefficiencies introduced by inaccuracies in the estimation of the products utility. I certainly am not arguing that it would be efficient to set the market value of all intellectual property to zero; after the initial burst of free, valuable IP created by the sudden change in the operating environment, it would quickly destroy the associated industries and the source of new IP would dry up to a trickle. Though IP protections have their own set of problems, they are a functional solution to an important problem; that IP needs to be monetized for there to be new investments into it.
123
124The alternative I offer would create efficiencies far eclipsing both the naked "free information" approach, and the artificial-scarcity system we have today.
125
126Envision an agency, perhaps national, perhaps international, but publicly-funded, not for profit, and which shares our objective of maximising utility. Let's call this agency the "Intellectual Property Bank" or "IP bank" for short. The IP bank offers contracts to all holders of varying types of intellectual property, indiscriminantly, and for a fee that represents only the cost price of administering the contract. The nature of the contracts is that the IP holder must place a copy of their IP in escrow with the IP bank, and when the contract expires, The IP bank will release the IP to all citizens under its authority for free, with no strings attached. In exchange, for the duration of the contract, the agency will subsidise all sales of the IP for an amount inversely correlated to the duration of the contract.
127
128For example, say that a software developer is releasing a new version of its flagship product. Let's say they they normally charge $50 per unit, and that normally their sales profile for the first 6 years looks like (in millions of units): 15, 5, 4, 3, 2, 1. So in total they sell 30 million units at $50/unit, for a total of $1.5 billion
129
130But instead, they choose to sign a 3-year contract with the IP bank. This will reduce their sales considerably, as some customers will hold out for three years for the free version of the product, and they will lose all sales for years after the third year. But in exchange for their trouble, they will get a 100% subsidy on sales made in those first three years. Since the impending and eventual release of the free version has reduced their sales, their new sales profile looks like this (in millions of units): 11, 3, 1, 0 , 0, 0. So in total they sell 15 units - half as much as before - but because of the 100% subsidy, their revenue is the same as it was under the old model. But more importantly, after the third year the software becomes free, meaning that we have generated huge additional utility at no cost. That additional utility will be equal to the number of people for whom the sofware would generate less than $50 of utility, multiplied by the mean utility they would gain from it. And this additional utility was generated at no cost, because the total cost of all software sales was the same in both examples.
131
132This solution, like any solution, does come with its own set of problems. Firstly, determining the appropriate size of the subsidy for a given length of contract would be a potentially complicated process. It would require somewhat accurate predictions of how much the IP holder's sales will be reduced by entering into the contract. Otherwise, you run the risk of over-subsidising them, or making the contract insufficiently attractive to encourage their IP's entrance into the public domain. This might necessitate a small team of analysts for generating quotations for potential contracts on a case-by-case basis.
133And secondly, it creates the same problem that every subsidy in the history of the world has created - the potential for fabricating fake sales of the product, to exploit the subsidy for the benefit of the company. I am not familiar enough with the regulatory environment to know how prevalent this is, or how difficult it would be to stop. But surely, the resources spent on preventing this kind of activity would be comparable to the resources saved in the courts on a daily basis, as the massive acceleration of the public domain reduces how much private intellectual property there is to be infringed upon, by an order of magnitude.
134
135An agency like this would be a huge undertaking, but for sufficiently large populations, and as software increasingly becomes a central pillar of our lives, the potential benefits of realising its efficiencies would be enormous.
136
137ON LAND OWNERSHIP
138Assets can be broadly divided into a few classes, based on how they traditionally impart utility.
139
140Many things which are bought and sold are useful exclusively for consumption either by a consumer or an industrial process, and offer no utility from their possession except to permit that consumption. This includes things like labor, food, steel, and lumber. Choosing to whether to produce one of these things is an assessment of how cheaply it may be done, and how much utility its creation would provide to society (as approximated by market value). In this way, the proper incentives are in place for the appropriate amount of these things to be produced, while using minimal scarce resources.
141
142The second class of things are those which are "consumed" gradually over their operational lifetime. Given that intellectual property protections eventually expire, intellectual property falls into this class of things too. These things provide utility simply by being owned - though using them may shorten their operational lifetime faster than if they were owned and not used. This includes things like houses, cars, and electronics. Choosing whether to produce one of these things is an assessment of how cheaply it may be done, and how much utility it's creation would provide to society (as approximated by it's market value). Items which are part-way through their operational lifespan can be inspected by potential buyers, and it can be exchanged for a reduced price based on how much life it has left. Eventually, its operational lifespan will run out, or labor and parts must be invested in it to keep it in good working order. If the utility it provided over its operational lifetime, to it's various owners, was greater than the cost of the scarce resources that went into its production, then creating the house/car/phone/etc was an efficient decision - and market value is an approximation of this. In this way, the proper incentives are in place for the proper amount of THESE things to be produced, while using minimal scarce resources.
143
144The third class of things is a unique category. This class contains a single member: Land.
145Land, with few exceptions (see: the netherlands), cannot be created. Therefore the concept of a cost of production is not applicable to land. It also does not get destroyed by consumption, nor does it have an operational lifetime, nor does it deteriorate. Furthermore, with trivial exceptions such as clearing, flattening, and soil remineralisation, it is also not improved by its owner. Despite the fact that buildings are referred to as "improvements", when you put a building on a plot of land, you have invested in the building, not the land. When the building is knocked down in thirty years time, the land will return to it's original state, but the land value will be different. Why? Because the lands value is not a property of the land itself, and is almost entirely outside the control of the owner. 
146
147The lands value is a property of the lands surroundings. The landowner did not build the nearby cornerstore, the landowner did not create the nearby job opportunities, and the landowner does not compel the trains to visit the nearby station. Ownership of land represents a constant consumption of access to these nearby desirable things, as well as the sunlight and rainfall that land receives. Two years of access to parks and transportation is twice as valuable as one year of access to parks and transportation. Two years of rain and sunlight is twice as valuable as one year of rain and sunlight. The "cost" of land ownership, (Not cost of production, but opportunity costs inflicted by your monopolisation of that land), is a function of time. To buy land, and own it forever at no cost, is simply incompatible with the economic reality of what land IS. An asset which is fundamentally different in all regards to all other assets need be treated differently by the rules of the economy. These traits - the traits of not deteriorating, not being created, and being consumed temporally, must be reflected by the price structure of its consumption. If the price structure of its consumption does not accurately reflect the costs inflicted by its consumption, this disconnect creates inefficiency. And this inefficiency is not meaningfully distinct from the same inefficiency that causes the deaths of diabetics or a lifetime lost in traffic; it is simply more opaque, and consequently, more difficult to rally voters behind. As beneficiaries of this inefficiency, landowners at least partially escape the need to contribute to society in return for their consumption of its fruits. The absence of this reciprocal contribution is the price paid by the rest of society for this inefficiency.
148
149There is some cognizance of this in the world already. Land value taxes are universally acclaimed by economists, and are a good solution to the problem. Some countries even have land value taxes already implemented. The vast majority of developed nations have either a land value tax or a property tax - but they are rarely enough to negate land rents entirely - and unless land rents are fully negated, land ownership is still burdensome to the society to some degree. Property taxes work similarly to land value taxes, but they are one of the many policies I referred to earlier as a "flawed approximation of efficiency". 
150
151The un-earned nature of land rents is also not a new concept. As long ago as Adam Smith's "The Wealth of Nations", he said - “As soon as the land of any country has all become private property, the landlords, like all other men, love to reap where they never sowed and demand a rent even for its natural produce.”
152
153Clearly at some point, somebody saw landowners getting money for nothing, and decided that something should be done about it. But even if adam smith understood the underlying problem, those who came up with the idea of property taxes clearly did not. And that is simply because a building is unlike land - a building is just like any other good, in the respects that it is costly to build, provides utility over its finite lifetime, and deteriorates. The decision into whether a building should be invested in should be dictated by market forces, and any deviation from this, such as a property tax, creates inefficiency. A property tax is essentially a land value tax, strapped to a wealth tax which discriminates against the owners of buildings. The validity or efficacy of a wealth tax is outside my area of expertise, but if your society has one it should be applied uniformly and not arbitrarily target the owners of buildings.
154
155So un-earned rent is a problem. Property taxes are a bad solution to that problem, and land value taxes are a good solution to that problem. Land value taxes cause some portion of the rent to go towards the entities which are actually responsible for making that land valuable. They may go towards the construction and maintenance of civic infrastructure that is shared by all the tenants and owners in the area - including the consumers who make the commercial land valuable, the employers who make the residential land valuable, and the employees who turn the gears of it all. They may go to the military responsible for securing the land against outside threats and to the services like police, fire and rescue which ensure the safety of the people on that land. As these are the ones who are responsible for the value of the land, not the landowner (except by dint of the landowner falling into one of these other categories), these are the ones who should be the beneficiaries of any rent attributable to the value of the land.
156
157And the exact opposite is true for any house or other structure that might be built on that land. The police did not pour its foundations, and the nearby train station did not erect the frame. Likewise, nearby employment opportunities did not lay one brick on top of another, and the commercial opportunity presented by the population density did not tile the roof. The building stands because somebody chose to build it. For this reason, the buildings owner should be the sole beneficiary of any rent attributable to the value of buildings that sit atop the land. Any taxes on the value of buildings atop the land are inefficient, and arbitrarily disincentivise the creation of buildings in comparison to other investments. Land value taxes are preferable to property taxes in every conceivable way.
158
159Land value taxes are implemented in some places, but as it currently stands, the portion of rental yields attributable to the value of the land greatly exceeds the land value taxes in all these places. As long as this is true, landowners are collecting un-earned rent. This is a great driver of wealth inequality.
160
161It is telling that modern social darwinists simply assume that the existing implementation of capitalism selects the virtuous to be wealthy. Any self-respecting social darwinist should be much more interested in analysing the current implementation to determine how it should be modified in order to select the virtuous with the greatest precision. The implementation, and intensification of land value taxes is one such modification. I am sure the community of professional economists knows of many more; I only presume to speak in their place because their voice is seldon heard in public political discourse.
162
163ON ENTREPRENEURSHIP, AND WELFARE
164Many types of inequality which are touted by people on the far left as proof of the systems dysfunction, do not represent un-earned wealth gained by the exploitation of inefficiency. Jeff Bezos did not gain 150 billion dollars of wealth exclusively by collecting rent, or being the beneficiary of inefficient monopolies, or by externalising costs to innocent third parties. Certainly he did benefit from every one of these things, but for the most part, Jeff Bezos gained 150 billion dollars of wealth by providing utility to other members of society at less cost than they could've gotten elsewhere.
165
166Certainly, large corporations own land, and emit pollutants, and operate monopolies, and reap the un-earned benefits of all these inefficiencies. Where this is the case, efficiency should be restored if possible, and if it cannot, then its benefits should be distributed to the people. But entrepreneurship such as what Jeff Bezos has undertaken, provides massive value to society, and as such it should provide massive rewards. The matter of whether 150 billion dollars is an appropriate reward is not self-evident, and to claim that the possession of 150 billion dollars is intrinsically inappropriate is intellectually dishonest. Whether 150 billion dollars is an appropriate reward for the activities Jeff Bezos participated in, is predicated on what beneficial activities Jeff Bezos participated in, and whether there is an incentive structure that would have been able to elicit equally valuable contributions for less cost than 150 billion dollars.
167
168As luck would have it, such ways exist.
169
170At its heart, like every economic decision, the decision to start a business is a cost/benefit analysis. The personal benefits of self-employment are increased agency and independence, the potential to realise a vision, and the potential for immense financial gain. The costs are the time and money that must be invested into a business to get it off the ground. To make the necessary investments and not be met with success, may leave the entrepreneur and their family destitute, so the risk of that is a cost too. 
171
172As with any cost/benefit analysis, the benefits of success and the costs of failure are assigned some weight or multiplier based on how likely they are. 50% of businesses fail within the first 5 years, and 66% of businesses fail within the first 10*. On a long enough timeline, the rate of survival for all businesses drops to 0. However, during the time a business is alive, the owner is able to pay themselves a salary for their work running it. It's unreasonable to expect a business to last forever; For every potential entrepreneur, there will be some level of longevity that, once achieved, the entrepreneur would consider it to have been a worthwhile endeavour.
173
174*Citation: https://www.sba.gov/sites/default/files/Business-Survival.pdf
175
176Completely arbitrarily, let's say that longevity is 5 years. Not because that is accurate, but because the survival rate at 5 years is 50%, and that will make the maths very easy.
177
178So you have a 50% chance of some degree of success, and a 50% chance of failure. 
179The low end of success looks like paying yourself a salary for a while, a lower salary than you would have gotten in the workforce but a salary nonetheless. Meanwhile you were able to reap the non-financial benefits of self employment for that time, and eventually call it quits and liquidate your business when it becomes clear it will never be greatly profitable.
180The high end of success looks like $150 billion dollars.
181
182There are ways to tweak the intensity of success to make entrepreneurship more attractive. You can lower the corporate tax rate, but this will do next to nothing. The vast majority of businesses in the united states states are not corporations, and do not pay the corporate tax rate. Out of 30 million, there are 1.7 million C corporations - the type which pay the corporate tax rate. All other business owners just pay income tax on their profits, but there is a third option. In Australia, there is a separate "small business" tax rate. Due to the fact that your business winding up small is overwhelmingly more probable than your business winding up big, this type of tax benefit can have a non-trivial impact on the expected value for profitability of a new business. But the magnitude of success is still only one of three variables. Lets look at the other two.
183
184*Citation: https://taxfoundation.org/us-has-more-individually-owned-businesses-corporations/
185
186The second variable is the probability of success. There are ways that small businesses can be cost-effectively given a helping hand, and doing so is an extremely effective strategy for increasing success rates*. Companies nurtured in a business incubator have a survival rate of 87 percent, compared with 44 percent for unassisted enterprises. This kind of insane number is essentially just the result of a bit of free office space and some mentoring, during the businesses fragile beginnings. There are likely many more ways to increase the success rate of new businesses, but the point is already made, so we move on.
187
188*Citation: https://www.ozy.com/acumen/can-an-incubator-increase-your-companys-odds-of-success/79735 
189
190The third and final variable is the cost of failure. Degrees of failure, much like degrees of success, vary greatly. A low degree of failure perhaps looks something like taking a few months off your job to start a small business in an industry with no barrier to entry, and then realising you're not cut out for it. All you would lose in this situation is maybe $20,000 of missed earnings.
191
192A high degree of failure looks like gradually sinking your life savings into a business venture that shows just enough promise to lure further investment (humans are quite prone to sunk cost fallacy), but not enough to reach profitability.
193
194The way to lessen the cost of failure, is to have a social safety net. An analysis of entrepreneurial activity before and after the implementation of a public child health insurance initiative in the united states (Olds, 2016), said the following of the program: 
195
196"Using data from the Current Population Survey (CPS), I show that SCHIP reduced the number of households with uninsured children by 40%, consistent with earlier estimates (Sasso and Buchmueller, 2004; Bansak and Raphael, 2005a). I also show that SCHIP increased the self-employment rate among parents by 15%. These businesses are disproportionately more likely to be high-quality ventures: SCHIP increased the number of incorporated businesses by 36% and the share of household income derived from the business by 12%. This is driven by a 12% increase in firm birth rates and a 26% increase in the rate of incorporated firm formation, and I document only small and insignificant gains in firm survival. I also observe a large increase in labor supply as a result of the program, both on the extensive margin (significantly more wage earners in affected households) and the intensive margin (longer work-years as measured by weeks and hours employed)."
197
198*Citation: Olds, Gareth. Entrepreneurship and the Social Safety Net (2016) https://www.hbs.edu/faculty/Publication%20Files/16-144_d9ce8326-eeaa-4650-a8af-6ff03c3f7e77.pdf
199
200So not only did this addition to the social safety net bring more entrepreneurs out of the woodwork, but those entrepreneurs were on average, more effective entrepreneurs than those which existed before the augmentation of the safety net - as evidenced by the 26% increase in the rate of incorporation.
201
202Entrepreneurship brings massive economic benefits, and should be incentivised as much as possible. But there are more efficient ways to tilt the cost/benefit analysis of potential entrepreneurs in favor of starting businesses than low taxes. Tax cuts grow the economy, but so does spending. The question of "Is it worthwhile to raise taxes and increase spending?" is essentially asking "How long is a piece of string?". To ignore the case-by-case complexities of taxation and spending is nonsensical - the answer is not a fundamental universal law. Some taxes increase efficiency, and some taxes reduce it. Some spending is wasteful, while some spending pays for itself many times over. The intellectual laziness of disregarding these variables is all too common - this is something that I will analyze in more detail in a later chapter.
203
204ON EXTERNALITIES AND JUSTICE.
205A popular myth is that all government intervention in the economy creates inefficiency. The grain of truth that this lie is wrapped around is of course the fact that all taxes and subsidies create inefficiency if they aren't in response to an existing inefficiency. But, naturally occurring inefficiencies are common, and by far the most common inefficiency is the "externality".
206
207For my less economically inclined readers (I doubt many of you are left, but I applaud you for making it this far), an externality is defined as a cost or benefit that is borne by a party that did not participate in the decision that brought about that cost or benefit. For example, my neighbor did not consent to me running my blender, but they still suffer the noise pollution. A farmer downstream did not consent to the contamination of his river by a factory upstream, but they still suffer the death of their crops. Externalities can be positive too: somebody that fireproofs their own home reduces the risk to surrounding homes. A business that stations a security guard at its entrance also deters crime for other businesses nearby. In all of these cases, the other victims or beneficiaries were not party to the decision. They were "external", and certain costs or benefits of the decision were "externalised" and borne by them in place of the decision maker. Thus the name, "externality".
208
209We already dispatched anarcho-capitalism during the introduction. The need for the state is clear, in that it is required to prevent usurpation by other less desirable states. The necessity of laws, as well as police and courts to implement them, is rarely questioned. But after that is established, arbitrary lines appear thick and fast. Usually the first and most conservative line is that the state should only prohibit physical violence. Then theft is included. Then of course, there are dangerous activities which are not explicitly aggressive, so these must be put a stop to. Shared resources like power lines shouldn't be pillaged, and rivers probably shouldn't be contaminated either. Guns are dangerous so less people will get hurt if we don't have those. If a business refuses to serve a customer because of the colour of their skin, that's certainly in bad taste, and death threats can cause pretty serious harm, so lets throw those in there too. Certain parking spaces provide extremely high utility, so it would be unfair to let one person leave their car there for too long. Loud noises can cause headaches, and certain quiet noises that people make with their mouths can make people sad. People might be slightly more likely to get hit by a car if they cross the road at somewhere that isn't a designated road-crossing zone. And so on, and so forth.
210
211The line between chaos and nanny-state authoritarianism is fuzzy, and where to draw that line is perhaps the most contentious issue in political discourse. What I interpret to be the biggest problem with discussions about these issues, is that people approach them like there is some kind of underlying objective truth to be found, on a per-issue basis. I propose that there IS truth to be found, but it is rarely objective, and the questions above have different answers depending on context. If there is an objective truth, I believe it to be this: everything comes at a price, and prices fluctuate.
212
213Now we have reached the beating heart of my earlier claims. My thesis was this: "The role of a nation's government should be only to maximise the efficiency of its economy", and its corollary: "this applies even to domains where a market does not or cannot exist".
214
215I claimed that, through the lens of economic efficiency, we could reinvent criminal justice as we know it, with "some important differences". The reason we are able to do this, is because externalities are a type of inefficiency. What is criminal justice, if not the process of internalising costs which a criminal has inflicted, or externalised, on a non-consenting party?
216
217The lesson I want you teach you in this section, and the examples that follow, is this: There is no meaningful distinction betweeen criminal justice and good economic policy. 
218I say it once more, for emphasis: There is no meaningful distinction between criminal justice and good economic policy. I have no doubt that you are trying to draw lines between those two things in your mind right now, to prove me wrong. But ultimately, all crime and punishment is a process of internalising costs - the process of forcing people to wear the costs of their own actions when they try to inflict costs on others. Anywhere you find injustice is where you see this principle deviated from. Think of anywhere you have seen the law bring injustice. Here are a few injustices, if you can't think of any: Imprisoning people for the possession of drugs or alcohol, excessive mandatory minimum sentences, imprisoning people for debts that imprisonment prevents them from paying.
219
220Think about all of these injustices, one by one. For each of them, ask the following questions:
2211. What are the costs of the action that was criminalised?
2222. What are the benefits of the action that was criminalised?
2233. Who are the beneficiaries and victims of these benefits and costs?
224
225Subtract any costs or benefits to the person who commited the crime - a lack of consent is a prerequisite for inefficiency, and the criminal implicitly consents to the consequences of their actions.
226
227Now you have a net cost or net benefit to non-consenting parties. If the action provides a net benefit, throw the case out; it should be encouraged, not criminalised. If the action provides a net detriment, proceed to stage two. Stage two involves asking these questions:
228
2291. What would it look like, for the criminal to wear the cost of their own actions?
2302. What additional costs would be incurred by the process of internalising this cost (forcing the criminal to wear it)?
2313. What additional costs would be incurred by their continued presence in the community? (ie. is this crime a predictor for future criminality?)
232
233If the costs of the internalisation process exceed the costs of not doing so, internalising it is inefficient. This is the case with the debtors prison: Obviously its bad for people to not pay their debts, and failing to pay your debts inflicts costs on others. But the costs of imprisoning a debtor are extremely high - you remove them from the labor force, thereby exacerbating the initial problem. You must pay to keep them detained, you destroy social cohesion, you reduce their chances of returning to the labor force upon release, and its just all-around a really inefficient idea. Not every bad thing needs to be punished - I know that idea may offend some people. Imprisonment is only efficient if the expected value of future damages exceeds the cost of imprisoning the person. If the expected value of future damages is less, then imprisonment violates the principle of maximising efficiency.
234
235Which brings us to prohibition. Certainly drugs and alcohol have negative effects; at the very least they contribute to drunk- or drug-driving and therefore we could say that the expected value of future damages of a drug offender is non-zero. But, "non-zero" is a far cry from $33,274 USD per year; the average cost of imprisoning somebody for 1 year in the united states, according to the Vera institute*. I don't have a citation for the following claim, but I'm just going to say it anyway because I'm hoping its not contentious enough to need one: I think it extremely unlikely that ANY non-violent drug offender is likely to cause more than $33,274 USD of damages in a year that are directly attributable to marijuana consumption (that is, the same person would have caused $33,274 USD less damage if they were sober that year). Let alone the AVERAGE non-violent drug offender. Given this information, imprisonment as a punishment for drug possession is economically nonsensical. Not that popular support is a good predictor of good policy, but nonetheless public opinion agrees on this issue; 59% of americans, and 74% of australians (my own country of origin) support decriminalisation of marijuana*. I was not able to find any polling data on alcohol prohibition, but I assume it is an extremely unpopular idea, and this principle does not differ at all drug to drug; the drug only dictates the expected value of damages. 
236
237*Citations: https://www.vera.org/publications/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends-prison-spending
238https://theconversation.com/most-australians-support-decriminalising-cannabis-but-our-laws-lag-behind-99285
239https://www.usnews.com/news/politics/articles/2018-04-20/poll-majority-of-americans-support-legal-marijuana
240
241Still, we accepted (perhaps erroneously; I accept it for the sake of discussion) that drug use causes damages to non-consenting parties. So what should be done? What is the answer to, "What would it look like, for the criminal to wear the costs of their own actions?".
242
243If drug use, on average, causes damages to non-consenting parties, then the damages should be paid for by the two parties that consented to those damages, at the point of sale. Those parties are, obviously, the drug user and the drug manufacturer. If the average use of a drug causes damages, the amount of damages on a per-use basis should be calculated as precisely as possible, and a tax of that exact amount should be added to the sale of each unit of the drug. In this way, the costs that once were externalised to innocent third parties, are once again borne by the users - and the revenue may be used to prevent, mitigate or reverse those damages, or benefit the citizens in other ways.
244
245If the relative tax on drug use (compared to taxes on other products) is less than the costs inflicted by drug use, then every drug sale represents a blow to the society's net utility. Conversely, if the tax on drug use is more than the costs inflicted by drug use, this is discrimination against drug users. Perhaps discriminating against drug users is not something that concerns you very much because you do not like drug users. Need I remind you, however, that the scenario of recreational drug sales is not meaningfully distinct from the scenario of cars, or sugar, or industrial solvents. Internalising costs is the algorithm for restoring efficiency to ANY externality, on ANY action or product, not just drugs. To deviate from this algorithm because of a bias against drug users means you are intentionally allowing inefficiency to exist in order to pursue a vendetta against a certain category of people in your society. If you really wish to do that, I would first take a moment to consider how many categories you fall into, and what other people might be likely to pursue a vendetta that includes you, before deciding whether that is a precedent you are comfortable setting.
246
247Lets consider another type of crime, to see if our algorithm holds up. Let us see what our "optimally efficient" solution looks like when applied to the crime of theft. Theft is a nice clear-cut example, as everybody is more or less in agreement that theft needs to be illegal. Let's start by looking at the current treatment of theft, and talk about what inefficiencies this treatment introduces.
248
249In the united states, the fines for theft vary by state*, and disciplinary action, is very much at the judges discretion. This is tacit acknowledgement of the inadequacy of using crime categorisation as a major determinant of sentences. In the united states sentencing guideline, the "offense level" for theft is clumsily broken up into discrete numerical categories, based on a number of arbitrary distinctions. The following is an excerpt from the united states sentencing guidelines, regarding theft.
250
251https://www.criminaldefenselawyer.com/crime-penalties/federal/petty-theft.htm
252https://www.ussc.gov/sites/default/files/pdf/guidelines-manual/2016/GLMFull.pdf
253
254"(a) Base Offense Level:
255(1) 7, if (A) the defendant was convicted of an offense referenced to
256this guideline; and (B) that offense of conviction has a statutory
257maximum term of imprisonment of 20 years or more; or
258(2) 6, otherwise.
259(b) Specific Offense Characteristics
260(1) If the loss exceeded $6,500, increase the offense level as follows:
261 LOSS (APPLY THE GREATEST) INCREASE IN LEVEL
262(A) $6,500 or less no increase
263(B) More than $6,500 add 2
264(C) More than $15,000 add 4
265(D) More than $40,000 add 6
266(E) More than $95,000 add 8..."
267
268This particular table proceeds all the way to P, for a total of 16 totally arbitrary and equally rigid value categories.
269
270Then follow five additional pages of arbitrary-yet-rigid modifiers and categories which intensify the nature of the offense to varying degrees. Three of these modifiers, I will include below for analysis:
271
272"(3) If the offense involved a theft from the person of another, increase by 2 levels.
273(4) If the offense involved receiving stolen property, and the defendant was a person in the business of receiving and selling
274stolen property, increase by 2 levels.
275(5) If the offense involved theft of, damage to, destruction of, or
276trafficking in, property from a national cemetery or veterans’
277memorial, increase by 2 levels."
278
279Though the law's dedication to labelling and categorising every circumstance that could intensify the severity of a crime is nothing short of admirable, creating an exhaustive list of all the tells that may reveal that a criminal is more or less likely to reoffend or at what intensity, is clearly impossible. These categories above, and indeed all five pages of the categories, are another example of what I call the "flawed approximation of efficiency" that humanity has cobbled together heuristically over centuries. All of the categories above can be expressed in terms of cost internalisation and mitigation. If you are able to understand and use the underlying principle, heuristics are no longer needed, and including them in the process is likely to impede you from arriving at an optimally efficient solution. Not only will replacing heuristics with the underlying algorithm greatly simplify the code of law, it will leave less room for human participation in the sentencing process, protecting all parties from corruption and capriciousness on the part of that human element.
280
281So, what does our algorithm for finding optimal efficiency have to say about some of these modifiers from the theft sentencing guidelines?
282
283"(3) If the offense involved a theft from the person of another, increase by 2 levels."
284As we should probably expect of all the heuristics, this serves a worthwhile purpose. Stealing from someone elses person requires an outsized impulsiveness and audacity, compared to, say, theft from a vacant premises. A theft from anothers person is a predictor of the likelihood and intensity of future criminality - exactly how strong a predictor it is, could be established through a statistical analysis of recidivism by past perpetrators of this type of crime. Once the strength of the predictor is established, it could be used to adjust forecasts of the damage that would be caused by a perpetrators continued freedom. Then, the expected value of damages attributable to the perpetrators continued freedom could be compared to the costs and damages (including lost laboor) that would involved in their detention, and the influence (whether positive or negative) that detention would have on their recidivism rate. Then you would be left with a number that indicates the benefit of their imprisonment. If positive, lock them up, if negative, let them go.
285
286This process, while it may seem complicated, does not consider any more variables than the existing approach. The difference is that instead of adding a "2" to some arbitrary offense level, it offers an accurate representation of the importance of that variable.
287
288The greatest benefit by far of this approach, is that it is "alive". Sentencing guidelines are unchanging documents. They are revised, sure, but those revisions are intermittent, and they may not necessarily have a scientific basis anyway. If circumstances in the society change, the sentencing guidelines do not change with it. If you did something that gets you five years, you get five years. But what if the cost of imprisoning people doubles? What if that person is performing an important role in society, such as parenting or employment? What if they are not performing an important role in society? What if the prison systems effects on recidivism deteriorate? What if the prison systems effects on recidivism improves? The sentencing guidelines were written four years ago. What was a good idea then might be a bad idea now, if variables in the economy change. This problem does not exist if you have the algorithm in front of you, and can use fresh data for the calculation of every sentence. If your countries prison system improves recidivism rates and rehabilitates people, then prison sentences should be handed out like candy. If your countries prison system just takes soft criminals and turns them into hardened criminals, then prison sentences should only be used as an absolute last resort. The fact that this variable is ignored instead of constantly monitored, and factored into the calculations of every single sentencing decision, is a sign of a systemic ignorance of the importance of that variable to determining the most efficient sentences.
289
290But, imprisonment is not the only punishment a thief recieves. They also pay fines, which is excellent, since it is a like-for-like repayment of damages. But what should the fines be? Is there an optimally efficient amount to fine somebody for theft? If we view this problem as a problem of maximising utility in an economy, there certainly is.
291
292The principle behind cost internalisation is that when an individual decides whether or not to undertake an action, the expected net benefit of that action to the individual needs to be the same as the expected net benefit of that action to the society. Any time those things differ, you have created waste, or inefficiency. Actions which have a net benefit to the individual but a net cost to society is a category that our society has attempted over the years to clumsily categorise as "crime", of which theft is an example. But as we have discussed earlier, crime is just the most extreme and obvious type of inefficiency, and for the sake of maximising utility, crime is not meaningfully distinct from other types of inefficiency such as monopolies and non-criminal externalities like pollution. When calculating fines that should be levied as punishment for crimes, the fine that would be maximally efficient is a fine that perfectly matches the net damages of the action to non-consenting parties, divided by the probability of getting caught, plus processing costs - this is the algorithm that makes the expected value of damages to the decision-maker equal to the expected value of damages to the society. Every part of that algorithm is crucial to to its proper functioning, but when applied correctly it has fascinating yet strangely intuitive implications.
293
294I repeat the algorithm here in isolation, for your convenience: "The fine that is maximally efficient is a fine that perfectly matches the net damages of the action to non-consenting parties, divided by the probability of getting caught, plus processing costs".
295
296The first of these fascinating yet strangely intuitive implications I mentioned, is that if an action presents no risk of causing damages to non-consenting parties, it should not be punished under any circumstances. This might sound obvious, but it has implications for strange edge cases like the german cannibal who had gotten consent from his victim at every stage of the process to confirm he did in fact wish to be killed and eaten. Assuming we take the consent of his victim as valid, then superficially this would imply that he should face no consequences. However, the victim of murder is not the only party that incurs damages from it - his family did not consent to his murder, and so its within the realm of possibility that he need be held accountable for damages incurred to them by the loss of their family member. But it does, I think correctly, imply that the penalty for killing somebody with their consent should be an order of magnitude less than the penalty for killing somebody without. It furthermore implicitly legalises the killing of people who not only wish to die, but have approval to die from their associates; effectively meaning that voluntary euthanasia is implicitly legalised by the algorithm if their family and friends consent.
297
298I will reiterate the statement we're analysing here, just to refresh your awareness of it: "When calculating fines that should be levied as punishment for crimes, the fine that would be maximally efficient is a fine that perfectly matches the net damages of the crime to non-consenting parties, divided by the probability of getting caught, plus processing costs."
299The second fascinating but strangely intuitive implication of this is that in all cases where the perpetrator both turned themselves in, and pleads guilty (and thereby raises his probability of getting caught to 100%), the maximally efficient fine is exactly equal to the expected value of damages caused to non-consenting parties plus processing costs. This implies that if somebody smashes a store window, steals some stuff, and then shortly later thinks the better of it and turns themselves in, returns all the goods, co-operates fully with police and pleads guilty, then they should only be liable for the damages they caused and the costs related to their arrest and processing.
300
301This also has the equally fascinating implication that it opens the door for something that I would like to term "efficient crime". Imagine you are walking down the street late at night, and are passing two buildings. One of them, is on fire, but the fire is still young and the building could saved with early intervention. The other building has a big entrance foyer with glass double-doors, and behind those doors you can see a fire hose reel in the foyer. Under the current system, smashing the glass doors in order to access the fire hose reel to put out the fire is indubitably a crime of breaking and entering. Admittedly the current system has its human element to serve as a failsafe in situations like this, so they would likely not be prosecuted. But the algorithm needs no failsafe - if a citizen knows that it is codified in the law that he is immune to prosecution from destroying lesser property in order to protect greater property, they will be much more likely to act in that situation and put out the fire. This is admittedly an extreme example, but that's sort of why this is so impressive; the ability for a simple algorithm to elegantly handle extreme edge cases with no modifications or intervention indicates an underlying reliability of that algorithm.
302
303While both interesting and promising, these edge cases are of relative unimportance compared to the base case. The base case of theft is of course that somebody takes something that belongs to somebody else with no intent to turn themselves in or return it, and must be caught by the police. In this case what is required to align the thiefs perception of the crimes benefit to himself with the costs the crime inflicts on society, there is an additional variable. It is not enough to simply fine them for the exact cost of the damages, because they have a sub-100% chance of getting caught. This would cause the crime to seem net profitable in their cost/benefit analysis of choosing to perform that crime. The percentage of crimes of this type which are caught and punished needs to be calculated. Ideally it should be calculated as precisely as possible, but if there is to be error it should be on the side of greater fines, because there may be some variance within the population of thieves, such that more elusive thieves would not be adequately disincentivised, and would make a profit that is subsidised by less elusive thieves. This is not necessarily a problem, as the costs of this inefficiency would be borne only by thieves, and it would mean the disproportionate receipt of penalties should naturally push less elusive thieves out of the "market" over time. Admittedly, this pattern might become a problem if there was difficulty extracting compensation from these less competent thieves, due to them having no wealth. Something that would make this easier is the concept of a UBI, negative income tax, or negative tax rate threshold - since this would mean some of every citizens income comes from the state, and such it can always be withheld - ensuring even thieves with no wealth will still be able to be given financial penalties. The merits and drawbacks of UBI, negative income tax, or negative tax rate threshold policies are not something I will go into any great detail on in this thesis.
304
305For those who are having trouble picturing exactly what I'm talking about in their heads, I will provide an example.
306
307Somebody commits an act of theft. The rate at which thefts of this type are prosecuted successfully is 20%, and the total damages to non-consenting parties were $200, plus the costs of their processing. This sets the probability of them getting caught at 0.2, and the damages at $200, so the efficient financial penalty would be ($200/0.2 = $1000) plus the costs of their processing.
308
309Applying financial penalties is always preferable to prison sentences, as prison sentences are only efficient for crimes where the cost of the perpetrators continued freedom is high. It is worth noting that, according to this algorithm, many of the factors that make financial penalties less efficient, make prison sentences more efficient; so they can be though of as complementary types of punishment for different classes of crime.
310
311For example, financial penalties are ineffective if the probability of getting caught is extremely low, or financial compensation cannot be fully extracted from the offender (potentially because the costs to be compensated are extremely large). If the probability of getting caught is extremely low, or financial compensation cannot be fully extracted from the offender, this lowers the disincentive to commit future crimes. Inadequate disincentives to committing future crimes is a predictor of committing future crimes, and is a circumstance that would raise the costs predicted to be inflicted by the perpetrators freedom. This would make prison time a more efficient solution in such situations.
312
313It seems, from this examination, that using economic efficiency as the determinant for sentences in criminal justice is an extremely effective strategy. However, it is not without its problems. There is, from what I can tell, exactly one problem. It is not a small problem, but I hope that it is smaller than the sum of the problems presented by the existing system, which seem endless.
314
315The problem with this approach is simply the difficulty in accurately calculating and forecasting damages. Where the expected value of damages of a course of action are in doubt, it will not be possible to accurately determine the most efficient punishment. This is most problematic when the "damages" involved cannot be quantified because there is no way for these damages to be compared to something with a market value, via the expression of preference. What exactly are the damages of seeing a couple having intercourse on the sidewalk in broad daylight? There is no part of the market that we can run a statistical analysis on to determine how much people are willing to sacrifice in order to prevent such a thing coming up, because there's effectively no aspect of personal choice involved in its probability. I do not see any way to implement the algorithm for offences like these. It may be unavoidable that the penalty must be decided on democratically. Despite the problems with democracy, it still seems a more sensible approach than it being decided by a small elite of people who haven't made any effort to sample the populations aversion to public copulation.
316In any case, there is always the possibility of falling back on the current system, where damages cannot be calculated or forecast for whatever reason. I would argue that even when the "what some dudes think" approach to sentence calculation does have to be employed as a last resort, it's efficacy will be enhanced if those dudes are aware of the underlying process that they are supposed to be emulating - that is, the process of maximising utility via market efficiency.
317
318I am yet to find a crime where the application of this algorithm does not give a satisfactory result. I challenge the reader to apply the efficiency-maximisation algorithm to arbitrary crimes and consider the results, but I will quickly cover some more cases I have considered in passing:
319
320The algorithm elevates the penalty for rape when the victim is underaged, due to the higher probability and intensity of psychological damage. Meanwhile it implicitly decriminalises consensual sex between similarly-aged couples that happen to straddle opposite sides of the 18-year barrier, since encounters of that kind are not known to cause damages (even if the definition of consent still caused that incident to be tried as rape, the penalty calculation would come up trivial).
321The algorithm holds employers accountable for failing to take precautions which would make workplaces safer, but does not compel them to take precautious which would be prohibitively expensive or onerous (for this purpose human life would have to be given a dollar value, but this is something that many institutions already do discreetly).
322The algorithm holds food producers accountable for failing to correctly label food. A customer only consents to the consumption of peanuts if the item was labelled as containing nuts. If peanuts are present in the item but not on the label, the producer becomes responsible for the resulting damages.
323The algorithm will hold people equally accountable financially whether the damages are intentional or unintentional - but if the damages were unintentional, you will be an order of magnitude less likely to get prison time, because intent would be a predictor of future criminality. This also seems apt.
324And finally, any case of repeated offenses will drastically increase the probability of getting prison time, since it will raise the expected value of damages to be caused by the perpetrators freedom. On the whole, one could expect one-off offenders to be much less likely to get time (due to the absence of minimum sentences), and that repeat offenders would be much much more likely to get prison time, and for longer.
325
326I think this is enough examples that I can leave it here. Though I have tried, I have truly been unable to find a case where the punishment generated by the algorithm seems unjust. I hypothesise that this algorithm is the one that our minds approximate when they calculate what is "just". By more closely describing the underlying objective of criminal justice - the process of internalising externalities, which are a type of economic inefficiency - this algorithm can be used to derive a new system of criminal justice which looks very similar to our own, but is markedly less irrational and prone to error. Perhaps it would be premature to place so much confidence in this approach that we rely on it primarily, but at the very least, it seems that anywhere that the current implementation of justice deviates strongly from this approach is worth looking at more closely. Because that is where we are most likely to find the law shirking its duties or overstepping its bounds.
327
328I said during the introduction that I was going to ignore a specific complication, but its impact on criminal justice is too important not to discuss briefly before moving on. That complication is moneys diminished utility (on a per-dollar basis) to the wealthy. The fact that this thesis measures utility exclusively in dollars without adjusting for the dollars varying utility on a person to person basis, raises the spectre of extremely wealthy people willingly paying a "crime fee" to do harmful things, at little "cost" to themselves, due to their indifference to the financial penalties. This is the most important example of an area where moneys varied utility to varying people means that disincentives calculated on a purely dollar basis will be an insufficient deterrent. This suggests that, though I have chosen to ignore this implication for the simplicity of this thesis, it may be unwise to ignore it in practice, unless the consequences of the crime are purely economic (and therefore those consequences might be fully reversed by the financial penalty).
329
330One aspect of crime and punishment that is incredibly contentious at the time of writing this, and therefore perhaps will be the greatest test, is the domain of the freedom and criminalisation of speech.
331
332ON SPEECH
333
334As I am on the the libertarian end of the political spectrum, I would classify myself as somebody who is substantially pro-free-speech. There are some people who classify themselves as "Free speech absolutists" however, and this is a term I have trouble understanding. If somebody who you have never met, publicly accuses you of a crime you did not commit, should that be allowed? Perhaps they think it should, but I find that hard to believe. Accusation is purely speech, and abolutism permits no exceptions. If there are exceptions, then free speech is not absolute. So the question of free speech is, much like taxation and spending, not a for or against matter, if you're interested in maximising utility. If you're interested in wellbeing, rather than enforcing some deontological morality, to simplify it to this binary choice of "for" and "against" free speech is incoherent. The question is once again, "where should the line be drawn?". Let us see what the damages vs enforcement costs algorithm has to say about this matter.
335
336In the most extreme case of the least acceptable speech, where somebody is wrongly accused of a crime, the outcome is obvious. The accuser is putting somebody in danger of the penalties that may be imposed by the law, and is also certainly damaging their reputation, and so must pay severe damages. In this case, the algorithm elegantly distinguishes between simply tarnishing somebody's reputation in a public forum, and prosecuting with false testimony, because these different actions generate different amounts of damages. Prosecuting with false testimony includes all the damages of tarnishing their reputation, plus the potential damages that might have been inflicted on them by conviction, and the certain damages that have been inflicted on them by their necessary involvement in their own defense. The additional audacity required to prosecute with false testimony might also inform the decision of whether they need to be imprisoned.
337
338On the opposite end of the spectrum, in the most extreme case of completely unobjectionable speech, there are obviously no damages.
339
340But where, exactly, does the algorithm draw the line that lies somewhere between these two extremes?
341
342There is a principle that has, this whole time, been underlying the process of economic entities making efficient decisions. It was not relevant, so I did not mention it, but it concerns the agent, and is eminently relevant to the discussion of free speech. That principle, is that agents must be well-informed in order to make efficient decisions. 
343If some speech is false, it will possibly cause inefficiency by impeding the abilty of other intelligent agents to make efficient decisions.
344If some speech is true, it will possibly increase efficiency by improving the ability of other intelligent agents to make efficient decisions.
345
346Therefore, by default, there should be no enforcement action taken against the utterance of truths - the enforcement costs of punishing true speech would be extremely high, since it would discourage the utterance of such truths in future and therefore pervasively deteriorate the efficiency of the entire system. Discouraging the utterance of falsehoods, on the other hand, does not reduce future efficiency and would in fact generally increase it, because incorrect information impedes the ability of intelligent agents to make efficient decisions.
347
348A good example of this is false advertising. Marketing is not an unproductive industry - it actually has the ability to play quite an important role in maximising utility. Marketing's legitimate role in an economy is to educate consumers about the product landscape, to help consumers match themselves to the product that will provide them the highest utility. This is a bit of a disconnect from what marketing often is today; an ongoing attempt to reverse engineer humanity's basest instincts in order to influence purchasing decisions that might be perfectly sound without that influence. Even this though, is nothing compared to the inefficiency introduced by marketing that provides false or misleading claims. In this situation, the customer is quite likely to get a markedly less useful product per dollar spent than if they had not been influenced by the false or misleading claim. This is something the world is thoroughly abreast of, and false advertising is already illegal more or less everywhere. Rightfully so.
349
350Unlike false or misleading claims, truths by default have a positive influence on efficiency, due to other intelligent agents requiring information to help them make predictions about the future. There are cases, however, where truths might be efficient to suppress; if there is extremely high utility in doing so.
351
352A few examples of efficient truths to suppress come to mind. The first that comes to mind is the suppression of a persecuted persons location. Though this information is "true", there is extremely high utility to be gained from it being suppressed, as its utterance may lead to that person coming to great harm. Therefore it would be efficient to suppress the truth of a persecuted persons location.
353The second example is essentially, "any truth which you are under contract not to disclose". This may sound like an extremely lazy definition, and it most certainly is lazy, but it's also the best way I can think of to describe the category. The signature of a nondisclosure agreement is essentially a record or proof, made in advance, of the utility an entity finds in the suppression of a piece of information. Much like the "efficient crime" phenomenon of the last chapter, it would be possible in theory to generate enough utility from the breach of a nondisclosure agreement that it exceeds the value of damages laid out in the nondisclosure agreement. But this is analogous to a traditional nondisclosure agreement stating fixed monetary compensation for a breach, and paying the compensation in full with the proceeds of the disclosure. 
354
355Other truths which you are under contract not to disclose might include speech restrictions in the terms of service of a social network, but the permissiblity of including such restrictions in a license agreement would decrease as a platform increases in size, since for large platforms this kind of speech suppression would constitute monopoly-enabled inefficiency. We talked about the gross inefficiency of monopolies in the chapter on patents, and the handling of monopolies will be covered in more detail in a later chapter. The short version is that this kind of speech suppression would be perfectly acceptable on a small platform, with a small market share, but on a monopolistic platform it may be grounds for the monopoly to be broken up. The monopoly being "broken up" does not necessarily mean the company must be broken up into two companies, or even the product broken up into two products. If the product can be bifurcated into two versions (one of which is efficient, and the other is inefficient), the monopoly would be broken. If we take twitter for example, this could be something like a flag on every users profile, which turns speech suppression on or off. The two versions of the product would then be "competing" for users, and twitter would be allowed to continue suppressing speech on the feeds of users who have opted-in to speech suppression, but they would not be allowed to do so for users who have opted out of speech suppression.
356
357There are likely more categories of truths that are efficient to suppress which I have not considered, than just these two. Information, and its value, are so highly variable that I suspect I would have no chance of ever generating an exhaustive list. It is enough to me that I have drawn your awareness to this category, and made some examples.
358
359Death threats are a very strange edge case, but I think the approach of maximising efficiency handles them elegantly.
360A death threat can be either true, or false. If its "true" (that they infact intend to kill), then it is extremely efficient to imprison the person who issued it, and infinitely moreso to imprison them before they can carry it out. If it is false (insofar as such a thing is knowable), then it is not efficient to imprison them, but they will be liable for any damages caused by the utterance of the falsehood. If we could know with 100% certainty that a death threat is false, and they never intended to physically harm anybody, then imprisonment would be totally inefficient. However, in the wild we must always deal with imperfect information, and instead must do our best to forecast the future damages of the perpetrators freedom - and whether they are imprisoned or not will always hinge on this case-by-case forecast.
361
362As for information which simply causes emotional damage, this is a highly variable situation. If you are in a public forum that has a code of conduct or terms of service, then you are implicitly consenting to hear any speech permissible by its terms of service or code of conduct, et cetera. So even if there are damages, they are not to non-consenting parties, and cannot be prosecuted. In a public forum with no conditions of entry however, then much like the earlier example of public copulation, the exact "code of conduct" for behaviour in a public space may unfortunately need to be decided on democratically or via the current system - our algorithm simply cannot make meaningful conclusions about what behaviours should have what penalties when there is neither a contract nor any way to establish a hierarchy of preference.
363
364Keep in mind that just because you are free to cause damages with speech to others in a public forum because they consented to participate in that forum, that's very different from something such as inciting violence in a public forum. The damages of incitement of violence are not caused by the speech directly, so consenting to hear the speech is irrelevant. The speech instead elevates the risk of damages by secondary agents. You consented to hear the incitement to violence; you did not however, consent to have violence performed on you. If the incitement is of a nature that it increased your probability of suffering violence, there would be a nonzero expected value of damages to non-consenting parties. This would also imply that the magnitude of this offense scales with reach. If we assume your average act of violence inflicts a utility reduction of $10,000, and the probability of a person who hears the incitement, ommitting that violence is 1 in 100,000, then damages would be $100 in the case where somebody with 1000 twitter followers incited violence. Damages would instead be $100,000 if somebody with a million twitter followers incited violence, and so forth. With greater power comes greater responsibility, and the variability of damages causes the algorithm to reflect this without special consideration.
365
366In a private forum, such as in person, private messages, etc. It might be theoretically possible that one could prosecute for emotional damages, but it would require several exceptional types of evidence, and the nature of this evidence ensures that only the most pervasive and malicious harassment will ever sum to non-trivial damages.
367
368Do you recall the introduction, where we spoke of markets? It was said that the only solution that can reliably extract an honest assessment of utility from a human being, is to determine how much they are willing to sacrifice in exchange for that utility. Thereby it follows that, the only way we can get an honest assessment of the utility of not hearing another persons words, is to determine how much somebody is willing to sacrifice in order to isolate themselves from that persons words.
369For example, if somebody says something via instant message, and you do not block them, you have taken no steps to ensure you do not hear their words, and we must assume damages to be zero. 
370If somebody says something, and you accept their phone call the next day, you have taken no steps to ensure you do not hear their words, and we must assume damages to be zero.
371If somebody says something to you in person, and in response you choose to leave that physical location ahead of schedule, and we can assume the damages to be greater than the cost of leaving that situation - so in this situation there are non-zero damages. However, the costs of leaving a physical location are extremely low, so for damages that small, it would still be inefficient to incur the costs involved in prosecution. The plaintiff (as the agent which decided to prosecute) would have to make up any shortfall between damages and court costs, which means that while you could successfully prosecute for damages this small, the plaintiff would have to pay more in court costs than they would recieve from the defendant. Also, all of the steps which must be taken in order to prove damages, serve to mitigate further damages, meaning that the chances of a case arising where the damages exceed court costs, are extremely low.
372
373So, all these factors conspire to make it incredibly difficult to prosecute emotional damages from speech in a private setting, but it is not impossible. The following conditions would have to be met: One would have to have taken a great many steps to avoid further contact with an individual, and have had these measures circumvented, until costlier and costlier measures had to be implemented. This situation is closely analogous to what is currently referred to as "harassment" in modern criminal codes.
374
375In the case of exceptionally fragile individuals (perhaps 98th-percentile and above) it is important to keep in mind that it is more efficient to isolate highly fragile individuals, than to isolate moderately triggering stimuli, because generators of moderately triggering stimuli far outnumber highly fragile individuals. This is similar to the story of the king who wished to have leather roads - and instead, one of his advisers showed him that he can achieve the same goal by covering only his feet with leather. So for a successful prosecution of emotional damage, the stimulus still needs to be above a certain level of severity that is independent of the emotional fragility of the victim. Specifically, the person must not be so fragile that the isolation of all people of that fragility or greater, would be more efficient than the censorship of all stimulus at least as triggering as the offending stimulus. A comparison of the stimulus to the damages it would inflict on an "average" person may be something that is difficult to measure. Its quite possible this is a function that must be served by judge or jury, an unfortunate reintroduction of the human element to the equation. However, it is probably still better to distribute that kind of power across all such decision makers than to hardcode a number into an unchanging piece of legislation.
376
377Opinions on free speech are widely varied, to the point that I do not think this chapter will make me any friends. However, regardless of whatever backlash I may recieve (from both free speech absolutists, and from advocates of hate speech laws; which would be struck down by this approach) I am nonetheless pleased with the results that were generated, and would happily live under a society with these laws. Unlike for conventional criminal justice, the system created by this approach is extremely similar to the laws we already have today - though there are certain countries with bizarre suppressive laws that would be struck down by this approach.
378
379One matter remains unresolved on the matter of speech. And that is, the matter of monopolies. I claimed that privately owned platforms would be free to censor so long as they are small, and that once they grow large enough, they need be prevented from abusing that position to create inefficiency - and that one way of doing that would be to force them to compete with themselves by bifurcating the platform. If you're not familiar with the problems with monopolies, this must seem like a very handwavey type of exceptionalism. But the problem is not that a private company is censoring its platform - they are free to alter their platform as they wish, so long as they are subject to the forces of competition. Whenever a company ceases to be subject to the forces of competition, its efficiency must be enforced artificially. This kind of artifice will never be as effective as true competition, which is why breaking up companies into smaller companies is the preferred way of dealing with monopolies. There are situations where this is not possible however, and those situations must also be dealt with. I have already railed against the destructive inefficiency of monopolies in the chapter on patents, but patents are just one type of monopoly, created artificially by the state. Monopolies also arise naturally, and are comparably inefficient in this form.
380
381ON MONOPOLIES
382
383In the event a monopoly comes to exist, it will create inefficiency so long as it is permitted to do so, in the manner described in the section on patents. This matter is something that can be addressed in a number of ways, and I will lay them out in what I understand to be the descending order of desirability, in this chapter. Its quite likely there are more monopoly disruption techniques that I am not aware of, but the inclusion of this chapter is not so much for academic progress, as it is for the raising of awareness.
384
385The most desirable way to disrupt monopoly is to introduce competition. This has long been the preferred approach going all the way back to the birth of antitrust law in 1890 and 1914, when the Sherman and Clayton acts were passed in the united states, respectively. It is by far the most efficient way to disrupt a monopoly, as returning competition to the equation essentially constitutes a full reversion to a naturally efficient market situation. This is not always possible however, as certain infrastructure may be prohibitively expensive to recreate; so much so that it would be inefficient to duplicate that infrastructure for the purposes of creating competition.
386
387In these cases it may be necessary to simulate the effects of competition; this might take the form of platform bifurcation like what was talked about in the twitter example, where two versions of the products frontend are forced to compete with eachother via a user preference. Or it might take the form of a limitation on gross profit margins, which would reduce costs to consumers in a similar way to what competition would do, while still permitting material ROI for the owners. A gross profit margin cap would be most relevant to infrastructure monopolies which are funded by sales to consumers, such as telecommunications companies and operators of tollways.
388
389Finally, failing any way to create competition, or to simulate competition, there may be no option but to redistribute whatever percentage of profits are attributable to the exploitation of the monopoly. This is by far the least desirable option, since the continued existence of the monopoly still generates considerable inefficiency when its profits are redistributed. This simply causes the un-earned benefits of that inefficiency to be spread across the whole population, rather than going to the owner of the monopoly. This is better than nothing, but it is little better than nothing - restoring efficiency is of far greater benefit to the populace.
390
391It seems certain to me that governments should greatly concern themselves with the matter of combating monopolies. The best techniques for classifying, detecting, disrupting, and mitigating the damage caused by monopolies, are not something I claim to be an authority on. Wherever it conflicts with my analysis, the academic literature on these matters should be held in higher regard than my analysis.
392
393
394ON TAXATION AND SPENDING
395
396The legitimacy of taxation is something that certain libertarian factions call into question. This is dubious, however. What we consider "ownership" of land is in fact a set of rights pertaining to exclusive use of that land. For these rights, and subsequently the land itself, to have any value, they must be enforceable through strength of arms. A single landowner being a state unto themselves will not find it possible to prevent a larger state with an organised military from using "their" land if they so choose, and potentially displacing the owner.
397
398So the necessity of a security force is not in question. And any landowner which shirks their responsibility to fund that security force will find their land rights not enforced by it, to be usurped by the first security force that sees fit to do so - including the one that they formerly paid protection to. So the implied threat of force is used to prevent free riders from benefiting from the existence of the security force. But the alternative to this is not independence, it is dependence on another, potentially more oppressive regime. Truly, government is a protection racket, but it is by far the cheapest protection racket invented to date, which makes it preferable to the alternatives.
399
400If you accept that it is legitimate for your right to the exclusive use of your land to be protected by the state, and for force to be used to ensure the provision of that protection, then there is little distinction to be drawn between this and the enforcement of other rights of ownership. Most commonly, this is extended to include the protection of citizens property and person from one another, as well as outsiders. If you do not accept some additional premises, then defense and law enforcment is the full scope of the state. This is at least an internally consistent worldview; it is not "faulty", in the sense that it is does not contradict itself, nor reality, like so many worldviews do. I do have one major criticism of it, however: This approach is not in the best interests of the population.
401
402That should not be a contentious statement; its a truth that is baked right into the definitions. If you wish to create a state that serves the best interests of its population, that is literally just a rephrasing of "maximising utility". If you accept the premise that the role of the state is to look after the wellbeing of its population, then you do wish to solve the problem of utility maximisation, and this opens the door to a number of things. The first, is competition laws, or anti-trust, which we spoke about in the previous chapter. The second, is taxation and spending in areas where it is sufficiently efficient to do so.
403
404The best example of highly efficient taxation and spending is fire and rescue services; support for fire and rescue services is a good litmus test for whether a person not familiar with the concept, believes that utility-maximisation justifies additional taxation.
405
406Fires are extremely destructive, and compared to the costs inflicted by fires, their suppression is quite cheap. It is several orders of magnitude cheaper to pay for some jocks and a truck to sit around waiting for a fire, than it is to just let the fire run its course and rebuild all the properties afterwards. So having a fire service is incredibly efficient. However, if fire services had a voluntary subscription, it would create a free rider problem. Since anybody that opts out of the service would still have a high degree of protection so long as properties adjacent to their own are subscribed to the fire service, they benefit from the service without contributing to its upkeep. Furthermore, their opting out does not make it any cheaper to provide the service to the remaining participants, since fires originating on unsubscribed premises will easily spread to subscribed premises. This means that the cost of subscription to the fire service would increase with each household to opt out of the service in order to free-ride. The solution to this is, of course, to force people to pay a tax involuntarily to prevent free-riding, and use the proceeds of that tax to provide the fire service.
407
408The same principle underlies all legitimate taxation. You cannot opt out of receiving the benefits of a fire service, so you cannot be allowed to opt out of the costs of its provision. There is likely no spending as efficient as a fire service - maybe garbage collection is a close second - so any other examples I provide will seem gradually sketchier and sketchier as their efficiency becomes less and less apparent. But the principle that justifies other types of government spending is not different to the principle that underlies the fire service. If you cannot opt out of the economic benefits a program provides, you cannot be allowed to opt out of the costs of its provision. And if the benefits of a program exceed its costs, it should be provided.
409
410There are certain types of government spending which you CAN opt out of the benefits of, at least partially. If somebody wished to opt out of public health insurance, for example, it would be possible for the government to not fund their health services. This only constitutes a partial opt-out, however, since you cannot opt out of the benefits of higher labor-force participation caused by a healthier population, for example. So even if they should be allowed to opt-out of the lions share of the cost of the program (commensurate with their ability to opt out of the receipt of insurance payouts), they cannot be allowed to opt completely out of its costs of provision, since they cannot completely opt out of its economic benefits. Furthermore, providing the ability to opt out of a service adds a minimum level of bureaucracy that must be maintained in order to track and process the status of different individuals, and opt-in and opt-out requests. Therefore all people who choose to opt out of the national health insurance program must also be held accountable for the costs of the opt-out program. Whether the costs of upkeeping the opt-out program, and the cost of paying for the second-order benefits of the insurance scheme, add up to be greater than the cost of simply opting in, is a matter of speculation. But if the sum of the cost of the opt-out program and the second-order benefits did exceed the cost of opting in, it would be completely sensible to dissolve the opt-out program on the grounds that no rational agent would use it, as the opt-out would be more expensive. If this was not the case, and the opt-out was cheaper, then the opt-out program should be preserved for people who want the option to self-insure. The outcome of this matter will be highly sensitive to the context of the economy in which the matter is considered.
411
412All government spending must be considered through this lens of costs and benefits. There are benefits to the spending, and costs to the taxation, and it should only be forged ahead with when the costs outweigh the benefits. For example, on the matter of the social safety net, there is a certain level of safety net which is efficient, that is, it more than pays for itself. The social safety net is known for doing things like increase entrepreneurial activity (Olds, 2016), increasing child nutrition with its longstanding effects on IQ (Stumm, 2012), and decreasing child poverty with its longstanding effects on criminality and health later in life (Holzer, 2008 and OECD). Through its effects on all these variables, it's quite possible for a social safety net to pay for itself many times over in this way. Conversely, too large a social safety net can reduce labor force participation, thereby inflicting hidden costs that are in excess of the costs of the program. Any given alteration to the social safety net must consider the evidence of secondary economic effects, and calculate whether the alteration will have a net positive or negative economic effect, before humanitarian considerations are taken into account. In any case, the question of how large the social safety net should be is an area of extremely active study. I cannot offer specific prescriptions for this matter because it is my understanding that the matter is not settled, and may never be due to the worlds ever-changing nature. All I will say is this: Absolute opposition to the social safety net has no economic justification, because the economic benefits can easily exceed the costs of the programs, even for people who are not direct beneficiaries of those programs. Likewise, indefinite expansion of the welfare state is not economically justifiable either, it should go without saying, because inequality is the carrot which we dangle before those people we need to play their part in turning the cogs of prosperity. But, proponents of indefinite expansion of the welfare state usually do not lean on economics, and prefer to lean on humanitarianism. When it comes to humanitarian issues, proponents and opponents alike are rarely swayed with the use of words. For this reason, I will refrain from comment. Your opinion is your own, and so long as it does not conflict with the economic reality of the situation, you are entitled to it.
413
414*Citations: 
415
416Sophie von Stumm (2012). You are what you eat? Meal type, socio-economic status and cognitive ability in childhood. https://www.hungrymindlab.com/wp-content/uploads/2015/10/von-Stumm-2012-You-are-what-you-eat.pdf
417Olds, Gareth (2016). Entrepreneurship and the Social Safety Net. https://www.hbs.edu/faculty/Publication%20Files/16-144_d9ce8326-eeaa-4650-a8af-6ff03c3f7e77.pdf
418Holzer, H. J., Schanzenbach, D. W., & Duncan, G. J. (2008). The economic costs of childhood poverty in the United States. Journal of Children and Poverty, 14(1), 41-61. 
419https://www.scholars.northwestern.edu/en/publications/the-economic-costs-of-childhood-poverty-in-the-united-states
420https://www.oecd.org/education/school/48980282.pdf?fbclid=IwAR3SF2VlRETvHVIqqv1Dy44LWBInPIDPusdRIKsDABS_K0dWDUKs2_6-rlU
421
422CONCLUSION
423Looking at government as a tool for solving the problem of utility maximisation has far-reaching implications. Markets are a necessary component for giving subjective assessments of utility appropriate voice, but when left unsupervised, the incentives of markets quickly shift from provision of utility, towards acquisition of monopoly, deception, and externalisation of costs. With the appropriate supervision, these inefficiencies can be corrected. When we find and correct the final inefficiency, maybe capitalism can be celebrated as the meritocracy that we all wish it was - that many claim it to already be. But that day is far away yet, and at our current rate of progress, may never come. In any case, my analysis is too primitive to take us that far, but it nevertheless illuminates a path forward for policy changes that any proponent of meritocracy and utility maximisation should uphold.
424
425If we wish to approach total efficiency, we can do so by eliminating non-beneficial sources of revenue, reducing the waste caused by price distortions (both government-induced and those created by externalised costs), and facilitating the free flow of information and competition. To this end, I recommend the following policy changes:
4261. Patents in their current form, the form where a state-enforced monopoly is offered as a bounty for development of new technology, should be replaced. Their replacement should allow competition, but still reward the development of new technology proportionately to how well it is received by the market. The simplest example of a system like this would be to simply cause the penalty for infringement to be a royalty to the patent holder, rather than a criminal offense.
4272. All property taxes should be scrapped, and replaced with Land Value Taxes of a similar magnitude. Once this is done, the land value taxes should be increased by 0.2% every year until the value of the tax is comparable to the proportion of property rents attributable to the value of the land. This will essentially mean that landowners can only collect rental yields on the value of their buildings, not on the value of their land. This will drastically reduce the barrier to entry for being an owner-occupier, while markedly increasing the ongoing costs of occupying land for a long period of time. This would also be expected to increase commercial land utilisation (Foldvary, 2005) as low-productivity landowners would be forced to cede the land to higher-productivity owners.
428*Citations: Foldvary, F (2005). Geo-Rent: A Plea to Public Economists. Econ Journal Watch, Volume 2, Number 1, April 2005, page 124
4293a. If a universal basic income (UBI), or similar policy such as a negative income tax is introduced, it should be subtracted from the minimum wage.
4303b. If a UBI or similar policy is not introduced, the minimum wage should be calibrated to a target unemployment level, rather than being static. That target unemployment level would presumably be the federal reserve target of 4.5%. Where the lowering of the minimum wage causes full-time employees to descend into poverty, that should be addressed as a failure of the social safety net, rather than as a failure of the minimum wage.
4314. Prison sentences should be abolished as a form of punishment for non-violent drug offenders, and any crime for which the perpetrators continued freedom does not imply a high probability of future damages.
4325. All products and activities which externalise costs onto non-consenting third parties should have those costs quantified and then be subject to a tax directly proportionate to the expected value of those costs. This category mostly consists of things with negative health, criminal, or ecological impacts, including emissions and drugs.
4336. All subsidies should be ended, except those which correspond directly and proportionately to a known positive externality, or are necessary for geopolitical reasons. This includes ending fossil fuel subsidies, which are unforgivably foolish economic policy. This also includes ending green energy subsidies. The correct approach to internalising the costs of fossil fuel emissions is to quantify those costs and tax fossil emissions by that amount, as described in (5).
4347. All contraband should be legalised, so that it may be subject to the taxes described in (5).
4358. Monopolies, particularly regional monopolies on infrastructure, need to be persecuted more aggressively. Competition must be introduced where possible. Where inefficient to introduce due to excessive barriers to entry, competition must be simulated. Where neither of these things are feasible, the monopoly should be deprived of the fruits of its inefficiency, and that money should be used to serve the people instead.
4369. The entry of intellectual property into the public domain should be greatly accelerated, perhaps through an "IP bank" style contract system, but at the very least, the duration of post-mortem copyright protections should be reduced by a minimum of 30 years.
43710. Powers and laws not justifiable as a means to utility maximisation should be treated with extreme suspicion, as they are by definition not in the interests of the people.
438
439Un-earned rents on land values and various types of monopoly (including IP) constitute major drivers of the "trickle-up" effect that creates and perpetuates wealth inequality. Likewise, problems like climate change are easily described and addressed as cost internalisation problems. Many of the problems faced by our society, such as where the line should be drawn on free speech, the illegality of drug use, and the worrying expansion of the police state, become black and white when converted to the domain of economics, and treated as a utility maximisation problem.
440It is my hope that this approach has, and will continue to cast some light on these problems in our world. And I hope the presentation of these solutions will be enough to rally people behind their implementation.