Welcome to our second week of lectures. This is the theory lecture of the week. In this lecture, I will present a general introduction and discussion of decision making in organizations. In the lecture, I will relate various rational system views of organizations that tend to focus on administrative units or leaders of organizations. A simple example organizational decisions can be found in the image in the slide doc showing a decision tree right next to me. The choice here is whether to upload a picture or not onto my Coursera course. A variety of criteria apply and help us decide. Fortunately, this particular image of a decision tree was taken off of creative commons and freely viewed. So we're okay, nonetheless, it gives you an initial sense of what we mean by decision making and the kind of process people use to make some decisions within organizations. This week, we read some work by James G Marge concerning decision making. Jim is a faculty member here at Stanford who spent several decades studying actual decision making behavior in an organizations. In the work you read, he classifies types of organizational decisionmaking that helps situate our readings further. Particularly, the rational and natural classes of organizational depictions that Dick Scott's work described last week. Jim March describes two general classes of organizational decisionmaking, or logics of decision making, as he calls them. The first of the Logic of Consequence or rational choice theory. And the second is the Logic of Appropriateness. What Allison, Graham Allison might call organizational process models. And others refer to as, as perhaps sense making. The core distinction between these logics is that, one is concerned with choices and instrumental efforts, and the other one is concerned with rule following and interpretive activity. Both are intentional forms of behavior. The former entails means and rational action, and the latter entails value rationality, or duty driven behavior. Value rationality contents that regardless of the cost, or without attention to them We often make decisions out of duty to a rule, a principle or a particular identity. The rational actor model is essentially a model that follows the logic of consequence. For the rational actor approach, the first aspect is knowing your alternatives. Here a decision maker asks, what are the options available to me? Second, it's important to know the consequences of these alternatives. Here one asks, what happens if I take each option? Third, you have ordered preferences or ranked goals and objectives in terms of greater or lesser value. Here you weigh the value you gained or lost by taking each option. The fourth aspect of a rational decision process with a logic of consequence, concerns using a decision rule or choice process. Graham Ellison refers to this as an inference pattern. Here, the choice process is a rule by which an alternative is selected on the basis of its consequences for preferences or goals. Two decision rules are commonly discussed and reflect different notions of a rational actor. The first is really an ideally rational person, traditionally called Economic Man by its critics. This individual is typified by clarity and knowing and consistency of preferences and objectives. They're an ideal form of a rational actor. The second kind of actor is a boundedly rational person and is typified by ambiguity and uncertainty in knowing incomplete information and inconsistencies in their preferences and objectives. So we are, here we have an individual who is more like the real person we all know and experience. Let's start with a simple example of a rational choice and action just to see the difference. Very soon it's going to be the rainy season in California and we call this winter. It dips down to about 60 degrees Farenheit and it rains quite a bit. Because we're, we're so tough, we always worry about whether to bring an umbrella or not. So let's say on a particular day, there's a 40% chance of rain. And we have to decide between bringing an umbrella, or not bringing an umbrella. Now let's say on that day, we see certain costs and benefits in each scenario. And the scenarios are pretty simple. Not bringing the umbrella, and it doesn't rain. Yeah, I didn't have to carry it. That was a great decision. Not bringing an umbrella and it rains, well, I get wet and that's kind of a downer. If I bring an umbrella and it doesn't rain, I have to carry it to around all day and there is a cost to that. It kind of doesn't fit in my bag and it, it's kind of a trouble. When finally, bringing an umbrella and it does rain, I'm prepared that day and I stay dry. So let's use, use some values from ten to negative ten in in order to kind of depict this. Let's say that not bringing an umbrella and it not raining is a six out of ten. I'm happy not to have to carry my umbrella all day. If I don't bring an umbrella and it rains, well, then I get wet. And I find that I, that really terribly disconcerting to be wet all day right? I'm going to give that a negative ten. Really is, really upsetting on our scale. On the other hand, say I do bring an umbrella and there is no rain. Then I have to carry it around. That's kind of a discomfort to walk around with it in my hands. So, let's give that a negative five. And then if I bring an umbrella and it rains, then I'm kind of pleased with myself for being so well prepared. I give myself a plus eight. So, below, or next to me here is this kind of matrix a chart. And you see kind of a decision tree where the changes of rain versus no rain is 40% to 60%. And then each of the options is listed right and we want to calculate the expected utility of each kind of scenario. And here, all we have to do is multiply not bringing umbrella which is plus six, with the chance of no rain which is 60%. And from that, we get a second value of three point six. So if you have six times point six, it's three point six. That represents the expected utility of not bringing an umbrella if it doesn't rain. But, if I don't bring an umbrella and it rains, it's negative ten times 0.4 or 40%. And that equals negative four. Now if I add those two together the, of not bringing the umbrella in both cases, I get the net expected utility of not bringing umbrella. And if we add those two we see that it equals negative 0.4. Now if I go through this same kind of operating of the lower branch of this tree for bringing the umbrella in the case of raining or not rai, raining, I find the expected utility of 0.6. Now if I compare the two, it's clear that bringing the umbrella, given my preferences or my sense of cost and rewards for each outcome. Is better than not bringing an umbrella because I really don't want to get wet. Now let's do this for a more interesting case of dating. Many of you are single and perhaps looking for love. And what I'm going to do now is show you how the, the logic of consequence can perhaps work to your favor as you consider your options. So let's take them one a, one at a time here. Say you're wondering whether to ask someone out. And let's say they're pretty attractive so you've got a ten percent chance that they'd actually say yes. Right? And here are our options. You don't ask them out and they say no. That's pretty good. Right? You're not embarrassed. The other option is you don't ask them out. But they would've said yes. In that case you miss out on something quite interesting and wonderful. That's kind of a downer. A third, you do ask them out and they say no. That's kind of mortifying, right? That may be terrible. Then finally there's the last option which is, we do ask them out, and they say, yes. When that happens, it's quite gratifying. How would you value each of these options though, from a ten to a negative ten? Are you a high-interest, low-cost person, or a low-interest, low-cost person, meaning, you know, you ask people out all the time and you don't have much to lose. Or you a, high-cost person, here you see it as a risky kind of endeavor no matter what happens? Now, me personally I find it kind of mortifying to be rejected. So I'll go with high-cost. And so in this table you see here not asking someone else out and then them saying no, well hey, that's good for me. I save myself the trouble, I'm going to give it a plus two, okay? Not asking them out, And they would have said, yes, that's kind of a downer, that would have been, let's say a negative eight for me. It's pretty bad but not terrible. But then, asking them out and then saying no is just awful. You know, I, I feel terrible and miserable over that. So I give that a negative ten, and then finally me asking them out and then saying yes is a plus ten. And that couldn't have happened better. Best of all worlds right there. So if we go through the decision tree again, and we can predict the net utility of each option of asking someone out or not, and let's say they're super attractive, right? Like I said before, and our chances are pretty low at ten percent, right? If we go through the math again like before, where we don't ask them out and get a yes, that equals negative eight. Then we multiply that by the probability of yes at 0.10, or ten percent chance. As such, negative eight times 0.1 equals negative 0.8 expected utility. The opposite of don't asking them out and they reject you, has a positive utility of 1.8. So we have a net expected utility of not asking people out equal to one if we add those two together. So in contrast, if we actually ask attractive people out, then given the risk and the probability that they will say no and that we would be modified, we have an expected utility of negative eight. So the net expected utility of that is pretty low of asking them out. So that's pretty severe. So of course, as a result, we just avoid the whole effort altogether. We, we don't ask them out, right? That gives you two examples now of a very simple decision tree that people make in terms of common types of decisions. But you can extend this to organizations and their types of decisions as well. And their kinds of options for example if a company does x then a competitor client has a probability or reacting in a certain way. Later I'm going to discuss the Cuban missile crisis as an example of organizational decisions. In that clase, case, there are really clear choices and there are potential outcomes. And, there are values affixed to each outcome. That's going to bring us closer to a real-world organizational case. Now clearly there's a ton of ambiguity here. Weather reports aren't that accurate, plus, I really have little good evidence to go on to decide if someone might be receptive to being asked out or not. Thus far, the Rational Actor Model is an idealized model that assumes Herculean abilities of decision makers. In reality, most of us are boundedly rational. We're unclear on our preferences and goals. We have vague information on the consequences and vague probabilities in terms of what we think others will do. So this leads us to our next question about ambiguity.
