WEBVTT NOTE An Entrepreneurial Theory of Moral Hazard 1 00:00:00.000 --> 00:00:06.440 So, once again, I'm very grateful to be here now to present some of the work from my PhD dissertation, 2 00:00:06.440 --> 00:00:11.080 which is titled The Theory of Moral Hazard, Foundations and Extensions. 3 00:00:11.080 --> 00:00:16.680 The ideas that I'm talking about today are excerpted from the second chapter of that dissertation, 4 00:00:16.680 --> 00:00:21.760 in which I attempt to show that there are some distinctly Austrian ways 5 00:00:21.760 --> 00:00:28.160 in which to think about problems of perverse incentives and moral hazard problems in particular. 6 00:00:28.160 --> 00:00:43.160 So relatively quickly, I'll just go over what moral hazard is in general, what the mainstream economists make of it, and then what the limitations of the mainstream approach are, and then lastly, how the theory can be understood in light of the Austrian theory of the entrepreneur. 7 00:00:43.160 --> 00:00:56.160 So I think it's best to begin just by explaining broadly what moral hazard is. I think moral hazard is one of those terms that was mostly unknown outside of economic circles until recently. 8 00:00:56.160 --> 00:01:04.160 But with the events of the financial crisis and the bailouts, you begin to see it used more commonly in everyday language. 9 00:01:04.160 --> 00:01:08.160 Moral hazard is a variety of welfare problem. 10 00:01:08.160 --> 00:01:13.160 And you can't really define it in terms of just one individual. 11 00:01:13.160 --> 00:01:17.160 You really need more than one person to make the idea make sense, 12 00:01:17.160 --> 00:01:23.160 because it's only when you have more than one person that welfare conflicts can emerge. 13 00:01:23.160 --> 00:01:30.160 Now the topic of social interaction broadly begins with the possibility for peaceful and productive association 14 00:01:30.160 --> 00:01:33.160 and that possibility of course leads to the division of labor 15 00:01:33.160 --> 00:01:40.160 but the extension of the division of labor requires basically that at some point delegation will be necessary 16 00:01:40.160 --> 00:01:45.160 some sort of duties will have to be assigned to one individual by another 17 00:01:45.160 --> 00:01:49.160 and once you have delegation a type of agreement exists 18 00:01:49.160 --> 00:01:52.160 say between a principal and an agent 19 00:01:52.160 --> 00:02:06.160 Although it may not be a formal agreement, and simultaneously it also becomes possible with the agreement for either party to this contract to behave in ways that are undesirable to the other. 20 00:02:06.160 --> 00:02:16.160 And this can happen either within a firm, say between a manager and a laborer, or even just in simple exchange between buyers and sellers. 21 00:02:16.160 --> 00:02:21.800 So, there are many different ways to define moral hazard, but the definitions that summarize 22 00:02:21.800 --> 00:02:29.600 best revolve around the idea that when an individual does not bear the cost of his actions, 23 00:02:29.600 --> 00:02:35.340 especially when he can push the cost of his actions onto unwilling third parties, he has 24 00:02:35.340 --> 00:02:40.800 less incentive to take care to avoid negative outcomes. 25 00:02:40.800 --> 00:02:45.960 In the simplest sense, reducing the cost of something encourages you to do more of it, 26 00:02:45.960 --> 00:02:54.960 and what the Theory of Moral Hazards is concerned with are a series of ways in which this very basic cost principle results in negative outcomes. 27 00:02:54.960 --> 00:03:03.960 The standard example is from insurance. When your house is uninsured against fire, you have a very strong incentive to take care of it. 28 00:03:03.960 --> 00:03:14.960 Because if anything happens to it, the cost falls on you. But for instance, if you are fully insured against fire, you don't bear the entire cost if the house goes up in flames. 29 00:03:14.960 --> 00:03:25.960 And therefore, you have less incentive to take care, for example, not to have flammable materials in your house, not to have oily rags piled up in the garage, that kind of thing. 30 00:03:25.960 --> 00:03:41.960 And in fact, depending on the particular circumstances, the lack of an incentive to avoid an event may actually be an incentive to bring it about, as in the case where a man burns his own house down to collect the insurance money. 31 00:03:41.960 --> 00:03:54.960 But whether we speak in terms of simply taking less care or actually actively pursuing some kind of undesired outcome, the principle remains the same. 32 00:03:54.960 --> 00:04:03.960 Moral hazard as a specific concept appeared in the 19th century in the insurance industry where the term was first coined. 33 00:04:03.960 --> 00:04:10.480 and, in fact, insurance remains the most well-known and studied example of moral hazard, although, 34 00:04:10.480 --> 00:04:16.760 as I will mention, the principle is actually much broader than simple insurance contracts. 35 00:04:16.760 --> 00:04:21.560 As far as economics specifically is concerned, moral hazard entered the literature through 36 00:04:21.560 --> 00:04:27.360 the work of Frank Knight in his 1921 book, Risk Uncertainty and Profit, which, of course, 37 00:04:27.360 --> 00:04:33.120 includes discussions of the insurance industry and the economic rules that govern it. 38 00:04:33.120 --> 00:04:37.520 Knight was obviously also one of the great theorists of entrepreneurship and exercised 39 00:04:37.520 --> 00:04:42.480 a very important influence on Mises and the Austrian theory of the entrepreneur. So in 40 00:04:42.480 --> 00:04:47.840 a sense, it should not come as a surprise that there is going to be a way of looking 41 00:04:47.840 --> 00:04:55.480 at moral hazard problems that is both distinctly entrepreneurial and familiar to Austrians because 42 00:04:55.480 --> 00:05:00.440 of this influence of Knight. And although Knight does not really dwell on the problem 43 00:05:00.440 --> 00:05:05.040 of Moral Hazard, he does indicate that it falls within the scope of entrepreneurship, 44 00:05:05.040 --> 00:05:08.520 which is what essentially I will argue in a moment. 45 00:05:08.520 --> 00:05:14.120 Knight's role in bringing the idea of moral hazard into economics is also important because 46 00:05:14.120 --> 00:05:22.120 it highlights the fact that the idea of moral hazard, along with many of its attendant problems, 47 00:05:22.120 --> 00:05:29.200 such as delegation, asymmetric information, adverse selection, and so on, these concepts 48 00:05:29.200 --> 00:05:34.760 were not invented by the economic mainstream of the 60s and 70s, but had roots in much 49 00:05:34.760 --> 00:05:41.480 more Austrian-friendly and entrepreneurial ideas. 50 00:05:41.480 --> 00:05:45.460 But unfortunately, after Knight's promising start, the idea of moral hazard dropped out 51 00:05:45.460 --> 00:05:50.360 of the literature for some time, and Knight's uncertainty-based sort of entrepreneurial 52 00:05:50.360 --> 00:05:53.320 view was basically forgotten. 53 00:05:53.320 --> 00:05:58.720 Moral hazard doesn't appear much in the non-insurance literature until the early 60s. 54 00:05:58.720 --> 00:06:04.720 One economist began to talk about economic behavior explicitly and sometimes exclusively 55 00:06:04.720 --> 00:06:07.200 in terms of incentives. 56 00:06:07.200 --> 00:06:12.560 And over the course of the decade from about 1963 to the early 70s, the idea of moral 57 00:06:12.560 --> 00:06:17.560 hazard was gradually formalized and included in a series of developing literatures and 58 00:06:17.560 --> 00:06:25.120 economics on incentives, contracts, and especially the economics of information. 59 00:06:25.120 --> 00:06:30.280 Kenneth Arrow published a famous paper on the market for medical care in 1963 that inspired 60 00:06:30.280 --> 00:06:34.280 like decades of research into moral hazard and there were other very closely related 61 00:06:34.280 --> 00:06:39.960 studies as well like Akerlof's 1970 paper on the market for lemons which was the paper 62 00:06:39.960 --> 00:06:44.560 that formalized the concept of asymmetric information. 63 00:06:44.560 --> 00:06:48.840 So what was really only through this sort of literature that moral hazard has come to 64 00:06:48.840 --> 00:06:53.600 be known in economics and the predictable result is that economists thinking on moral 65 00:06:53.600 --> 00:07:06.600 The behavioral hazard is usually based on a few conventional assumptions about behavior that are common, that are actually matters, of course, for the mainstream, but are not really compatible with the Mungarian tradition. 66 00:07:06.600 --> 00:07:14.600 For instance, the idea that behavior should be modeled as deterministic as opposed to involving some kind of real choice. 67 00:07:14.600 --> 00:07:21.600 Second, that behavior involves consistent maximization of clearly defined objectives. 68 00:07:21.600 --> 00:07:32.600 Third, that the results of choice can be analyzed and welfare implications drawn in terms of the properties of some relevant equilibrium. 69 00:07:32.600 --> 00:07:44.600 Specifically, the outcome of decision-making under moral hazard can be defined and analyzed in terms of a comparison between different market models. 70 00:07:44.600 --> 00:07:59.600 Specifically, some well-defined market model where moral hazard exists compared to some equally well-defined model where there are no perverse incentives or where there is no moral hazard at all. 71 00:07:59.600 --> 00:08:05.600 Typically, some variety of perfectly competitive model. 72 00:08:05.600 --> 00:08:13.600 In practice, this means constructing models of moral hazard that start with a description of some kind of agency relationship. 73 00:08:13.600 --> 00:08:20.600 Typically, both principal and agent are fully rational utility maximizers, each with their own specific decision functions. 74 00:08:20.600 --> 00:08:30.600 The relevant incentives that push people in one direction or another are implied in their respective functions and in their faculty of rationality. 75 00:08:30.600 --> 00:08:38.600 As is conventional with this type of decision making, individuals are essentially at the mercy of the incentives they face. 76 00:08:38.600 --> 00:08:54.600 As Kirzner puts it more generally, a chosen course of action, because it was pronounced mathematically to have been the optimal course of action within the given decision framework, cannot fail to be chosen again and again so long as that given framework prevails. 77 00:08:54.600 --> 00:09:03.600 And from these decisions that people make, we can derive various properties relating to equilibrium from this given set of starting assumptions. 78 00:09:03.600 --> 00:09:17.600 And then once we have different outcomes to compare, we can then talk about welfare, and that in turn is supposed to tell us if moral hazard results in some kind of like welfare loss that we should be concerned with. 79 00:09:17.600 --> 00:09:32.600 The result of this reasoning, this mainstream approach, the reason why we're supposed to care about it, is that moral hazard is sometimes used to demonstrate the existence of market failure and provide a rationale for some type of intervention. 80 00:09:32.600 --> 00:09:44.600 The idea is that markets sometimes provide these sort of perverse incentives for people to behave in undesirable ways that lead to non-optimal welfare outcomes. 81 00:09:44.600 --> 00:09:53.600 So for example, if health insurance companies provide full coverage, they end up encouraging people to engage in all types of risky behavior. 82 00:09:53.600 --> 00:10:00.600 So essentially by trying to help people avoid a negative outcome, you end up subsidizing that outcome. 83 00:10:00.600 --> 00:10:08.000 So, you know, on the sort of negative results this is supposed to bring about, they come in a few different varieties. 84 00:10:08.000 --> 00:10:17.000 On the one hand, people can simply sort of abuse the system and consume more of a service than is deemed optimal. 85 00:10:17.000 --> 00:10:26.600 In the case of like a health insurance, they would be consuming more medical services than is deemed optimal by the insurance firm. 86 00:10:26.600 --> 00:10:56.600 But then on the other hand, perhaps more importantly, there's a broader sort of market implication, which is that, again, to take the insurance case, insurance companies will presumably know that this is how people behave when facing certain incentives within a certain type of contract, so insurance companies can simply respond to this expected behavior by not offering certain types of contracts that they believe will 87 00:10:56.600 --> 00:11:02.400 involved these incentives and so the end result in terms of the market for something like 88 00:11:02.400 --> 00:11:10.600 insurance will be that there are fewer, there's a smaller menu of contracts and simply fewer 89 00:11:10.600 --> 00:11:16.280 insurance services available for purchase. So the market for insurance will be smaller 90 00:11:16.280 --> 00:11:20.640 than it might conceivably be and presumably there will be people out there who are sort 91 00:11:20.640 --> 00:11:45.960 And again, this applies to many different types of markets, not just insurance, any time we would have these sorts of conflicting incentives at play, those could also be cases of moral hazard, so it should not come as a surprise that this is a rather mechanistic way of looking at things, especially in that there's really no room in these models for real choice or true 92 00:11:45.960 --> 00:11:53.960 There's really no room in these models for real choice or true uncertainty, and especially not for entrepreneurs. 93 00:11:53.960 --> 00:12:06.960 The outcome of this process is already implied in the conditions of the problem, in the decision functions of the principles of the agents or what have you. 94 00:12:06.960 --> 00:12:24.960 And in my opinion, trying to set the theory of moral hazard on firmer ground involves applying a series of just very simple Austrian insights to this sort of conventional view, particularly for allowing for true choice and especially for uncertainty. 95 00:12:24.960 --> 00:12:36.960 So just to see how Austrians might approach moral hazard problems, we can just think in terms of a conventional agency problem, where a principal is delegating some sort of responsibilities to his agent. 96 00:12:36.960 --> 00:12:46.960 Moral hazard is most commonly modeled as a principal agent problem, where the principal is an employer and the agent is a laborer or a subordinate of some sort. 97 00:12:46.960 --> 00:12:56.800 Now, the problem arises mostly with regard to information asymmetries. 98 00:12:56.800 --> 00:13:01.920 The problem is essentially that if the principal could monitor the agent at all times, there 99 00:13:01.920 --> 00:13:03.560 would be no problem. 100 00:13:03.560 --> 00:13:08.000 Because there would be no opportunity for hidden behavior, which is another expression 101 00:13:08.000 --> 00:13:10.240 for moral hazard. 102 00:13:10.240 --> 00:13:17.620 But because monitoring is imperfect, it becomes possible for the agent to get away with not 103 00:13:17.620 --> 00:13:24.100 doing his job all the time or putting in less effort than he agreed to. 104 00:13:24.100 --> 00:13:31.900 And by doing this, he imposes some kind of additional cost on the principle. 105 00:13:31.900 --> 00:13:37.140 In the conventional view, that would be basically it, end of story, because the agent must pursue 106 00:13:37.140 --> 00:13:41.060 to the incentives that define his decision function. 107 00:13:41.060 --> 00:13:44.820 But in the real world, of course, the problem is that just because the possibility for shirking 108 00:13:44.820 --> 00:13:48.460 exists doesn't mean that it will occur. 109 00:13:48.460 --> 00:13:53.520 People are not utility maximizers in this very restrictive sense, and they face an enormous 110 00:13:53.520 --> 00:13:59.180 range of incentives all the time, only one of which is going to be the incentive to, 111 00:13:59.180 --> 00:14:02.400 to say, slack off. 112 00:14:02.400 --> 00:14:05.680 No one is obliged to follow any particular incentive. 113 00:14:05.680 --> 00:14:08.360 You might, you might not. 114 00:14:08.360 --> 00:14:12.880 And this is just a very, this is an extremely simple idea, but I think it throws a lot of 115 00:14:12.880 --> 00:14:18.640 the idea of moral hazard into confusion, because by introducing the concept of real choice 116 00:14:18.640 --> 00:14:25.160 and action into an agent's decision, you can also introduce uncertainty about what that 117 00:14:25.160 --> 00:14:27.560 decision will be. 118 00:14:27.560 --> 00:14:31.760 And naturally, once we start talking about uncertainty, this is where the entrepreneur 119 00:14:31.760 --> 00:14:34.020 is going to come in. 120 00:14:34.020 --> 00:14:40.260 Because it ends up being one of the entrepreneurs' fundamental duties and tasks to judge and 121 00:14:40.260 --> 00:14:45.020 arrange the incentives that are present, say, within a firm. 122 00:14:45.020 --> 00:14:51.580 And more importantly, it's a task of the entrepreneur to judge how laborers or agents will react 123 00:14:51.580 --> 00:14:54.980 to prevailing incentives. 124 00:14:54.980 --> 00:14:56.920 Are workers likely to shirk? 125 00:14:56.920 --> 00:15:02.980 If so, how different will their performance be from the duties outlined in the labor contract? 126 00:15:02.980 --> 00:15:08.380 What sort of effect will that behavior have on overall productivity? 127 00:15:08.380 --> 00:15:13.900 These questions have to be addressed by the entrepreneur who speculates about the hidden 128 00:15:13.900 --> 00:15:19.220 behavior of his agents according to his own judgments. 129 00:15:19.220 --> 00:15:24.740 The entrepreneur as the residual controlling force in the production process plays the 130 00:15:24.740 --> 00:15:30.940 fundamental role in arranging incentives so as to promote the success of his own enterprise. 131 00:15:30.940 --> 00:15:36.100 And in practice, that will mean that the entrepreneur will basically make judgments about how the 132 00:15:36.100 --> 00:15:41.300 arrangement of incentives will affect the productivity of labor. 133 00:15:41.300 --> 00:15:46.220 And then further, he will also make judgments about how expected labor productivity will 134 00:15:46.220 --> 00:15:51.500 compare to expected productivity of capital and so on. 135 00:15:51.500 --> 00:15:55.460 And so the entrepreneur has to constantly account for the presence of hidden action, 136 00:15:55.460 --> 00:16:01.540 in information, in the incentives that, within his firm or within his market, and this is 137 00:16:01.540 --> 00:16:07.500 really just another component, in my opinion, of the notion of entrepreneurial calculation 138 00:16:07.500 --> 00:16:10.740 that Mises suggested. 139 00:16:10.740 --> 00:16:14.980 And I think that looking at things this way lets us contrast the conventional from the 140 00:16:14.980 --> 00:16:19.140 Austrian view, because when we acknowledge that arranging incentives is an entrepreneurial 141 00:16:19.140 --> 00:16:24.940 and Entrepreneurial Activity, we can see that the arrangement process is open-ended just 142 00:16:24.940 --> 00:16:30.020 like other entrepreneurial ventures. 143 00:16:30.020 --> 00:16:33.620 Entrepreneurs can be successful in figuring out how to properly motivate people and inspiring 144 00:16:33.620 --> 00:16:39.200 them not to shirk, but entrepreneurs can also fail and provide disastrous ways of remunerating 145 00:16:39.200 --> 00:16:43.300 agents such that they ruin the entrepreneur's own business. 146 00:16:43.300 --> 00:16:48.720 But the overall point is that there are always very broad arrays of incentives that are available 147 00:16:48.720 --> 00:16:55.800 and the interaction between principle and agent, especially between entrepreneurs and their 148 00:16:55.800 --> 00:17:04.720 subordinates is likewise a constantly shifting problem of anticipation and speculation. 149 00:17:04.720 --> 00:17:09.920 This entrepreneurial behavior is just a necessary part of the market process. 150 00:17:09.920 --> 00:17:17.400 It's not an aberration as it's often treated in the mainstream literature. 151 00:17:17.400 --> 00:17:21.060 So pointing out that incentives are a problem for entrepreneurs is one way in which moral 152 00:17:21.060 --> 00:17:25.340 hazard can be thought of from an Austrian perspective. 153 00:17:25.340 --> 00:17:29.520 Another very important way that I'm going to have to skip over, although actually it's 154 00:17:29.520 --> 00:17:33.460 fortunate because in a sense it's already been covered today, another very important 155 00:17:33.460 --> 00:17:39.760 aspect of these moral hazard problems isn't just defining what negative welfare outcomes 156 00:17:39.760 --> 00:17:42.340 are to begin with. 157 00:17:42.340 --> 00:17:47.820 This morning in the author's forum a couple of different people brought up this idea of 158 00:17:47.820 --> 00:17:51.620 the nirvana fallacy that you can't compare the real world to some sort of abstract equilibrium 159 00:17:51.620 --> 00:17:54.860 where all the relevant problems have been assumed away. 160 00:17:54.860 --> 00:18:00.540 This gets right at the core of the idea of moral hazard because in the mainstream literature 161 00:18:00.540 --> 00:18:04.820 it's constantly defined in terms of one of these equilibrium points. 162 00:18:04.820 --> 00:18:11.540 Unfortunately, I'm out of time so I won't have time to talk about any more explicitly 163 00:18:11.540 --> 00:18:41.540 Austrian takes on this, but essentially, I mean, I think you can anticipate the punchline of this. And that will be that true moral hazard problems are going to be emerging in situations in which this entrepreneurial market process is disturbed, particularly when it's disturbed through some type of coercion, be it private or public. As I said, there are many more details in this, but unfortunately, I'm out of time. So thank you for your attention. 164 00:18:41.540 --> 00:18:43.540 Thank you very much.