The Art of Decision-making
There is no easy decision in life. Every time we are facing a set of possibilities, each with a different outcome and with a certain degree of uncertainty, many factors come into place that make us hesitate about what should we choose. Think about when you had to decide which major to choose in college. It’s not just about what you liked the most at that time, it also involved factors like marketability, your set of skills, quality of the universities you were considering, etc. Every decision, as simple as it looks at first, comprises a complex analysis that takes into account several factors that most of the times are “in the back of our heads”. What I mean by this is that as intelligent creatures, most of the processes taking place at the moment of decision-making, which are all the mathematical and economic optimization that lies behind every decision we make, are ignored by our “conscious” brain. As an Economics student I am fascinated by how humans make decisions, and I believe that every decision in life, as complex as it could be, can be simplified into three easier decisions.
Decision-making in Economics used to be approached by the typical microeconomic perspective of simply maximizing what we call a utility function subject to a budget constraint. A pretty straightforward mathematical technique that anyone with a basic multivariate calculus knowledge could do. This basically sums up years and years of economic theory and it is the most basic approach to what motivates an individual to make a choice. You have some preferences, modeled by a “utility function” that has no value by itself more than ordering sets or bundles of goods. For example, if you like pizza more than broccoli, your utility function will give a greater weight to the amount of pizza you consume compared to the amount of broccoli. Pretty simple, right? There are some more details to take into consideration. Imagine you are in the desert, dying of dehydration, and you find a water fountain. The value of the first cup of water you drink will be enormous. The second cup of water will give you a great “utility” too, but not as much as the first one, and so on. This is the concept known in Economics as “Diminishing Marginal Utility”, which also has implications in the mathematical definition of a utility function. Decision-making is not only about preferences, though. It is also extremely important what you can and can’t afford, which is represented in your “Budget Constraint”. I might love Ferraris a million times more than I love Toyotas, but that doesn’t mean you will see me driving a Ferrari. So, your decisions theoretically depend on two things: preferences and money. Assuming we are all rational individuals, every time we make a decision we are doing this optimization process in our head, finding the “optimal consumption point”, something like this:
If you have never taken an introductory economics class (and even if you have) you are probably thinking: “What?… I’ve made decisions all my life and never had thought of anything even related to this”. And that is one of the main points I want to make. All this theory is just theory. It all lies on the assumption that people are rational, which implies that we are instinctively making this computation in our heads every time we face a choice. This is of course a hard assumption, and many researchers have dedicated their lives to prove that humans make “irrational” decisions. In order to talk about this type of behavior, I want to emphasize what the real foundations of decision-making are. To do that, I will decompose the process of decision-making into three straightforward trade-offs, that can summarize any decision that you make:
Now vs. Later: Almost all decisions imply a time-preference choice. How you feel about the future is extremely important when it comes to deciding over something. Would you prefer receiving 100 dollars right now or would you rather wait 3 months and receive 200 dollars? What about receiving 100 dollars in a year or 200 dollars in a year and three months? Tons of experiments have been done with questions like this, finding that some people are more focused on the present than others. In the typical economic theory, based on the Samuelson’s discounting model, the answer to these two questions should be the same, however, in real life the results are not always loyal to the theory. Every person values time differently. Some of us are more focused on the present while others don’t discount the future as much (A little bit more of theory here). You might be wondering “And how does this affect decision-making?”, in many and really important ways. A large part of economic literature studies the differences in time discounting depending on income, in other words, trying to look for a relationship between being poor and being more focused on the present (For example this). Think of it in this way: If you are poor, you are not thinking of investments that could pay off in five or ten years, you are thinking of having enough money to buy food this week. As a poor person, you are more vulnerable to shocks like unexpected expenses or the consequences of getting fired. The general conclusions support the idea that poor households are more present-oriented, and what are the consequences of this? Underinvesting in their children’s education, for example, which will make it harder for the children to get out of poverty. You might be thinking “Oh well, I am not poor, so this doesn’t necessarily apply to me”. Wrong. Remember that time you bought that 3-month gym membership thinking “I’m gonna to the gym everyday and get super fit” and ended up going 5 times? Yes, that is also a time-discounting problem. Before swiping your card and signing that deal that will make you a proud member of Gold’s Gym for three months, you have both the costs and benefits of working out in the future. However, when it finally comes the time of getting up from that couch, putting on your gym clothes, and going to lift some weights, the cost is immediate (think of it as the physical cost of going to the gym instead of doing something else), while the benefit (getting ripped, having a better “personal image”) is in the future, so if you are more present oriented, the immediate costs might outweigh the future benefits. One of the professors I had in Berkeley did a pretty nice study about this, that you can check out here (Another great paper here). Finally, one of the biggest issues nowadays regarding time discounting is smoking. This is the typical case of immediate benefits and future costs. If you are more present-biased, the pleasure you get from smoking a cigarette can outweigh the future costs that it will pass on your health. So does all of this implies that the rational model is wrong? What should be use to judge efficiency, rational models or observed behavior? There is no clear answer.
Self vs. Others: The second trade-off in any decision to be made is pretty simple: how much weight do you put on your personal benefits/costs against others’. Consider the following problem: Would you rather get $1000 and have a stranger (who you don’t know nor you’ll ever meet) receive nothing, or split the money so you both get $500? This type of decision, often times called distributional preferences, is a determinant factor in policymaking nowadays. Some people, for any given reason, could be more “fair-minded”, meaning that they prefer everyone having the same amount of money over keeping all the money, while others might prefer the opposite. There is no right or wrong answer, and that is the beauty of decision-making, because it is based on preferences (ethical, moral, economical, however you want to define them), so any answer is perfectly understandable. We can add another dimension to this problem, the big trade-off between equality and efficiency, and that is how we can end up talking about politics. How much should people be taxed? Would you rather have an efficient economic system that leads to high levels of inequality or an inefficient (in economic terms) model that makes everyone’s income equal? Your preferences in this matter are more important than you think. In theory, these preferences (and the aggregate preferences of the population) could determine voting outcomes, depending on the political parties’ economic plans. For those of you who find this interesting and would like to look into a more technical side, I attended a talk by Shachar Kariv, expert on behavioral economics and Chair of the Berkeley Economics Department, where he talked about a paper he had been working on that explains exactly this. Estimating Constant Elasticity of Substitution (CES) utility functions, he and his co-authors model distributional preferences and political behavior. You can find the paper here.
Risk vs. Rewards: Finally, the last trade-off present in all decision involves how much you are willing to take risks. You might be expecting a question like “Would you rather invest $X with certainty or $Y with uncertainty?” in order to make my point, right? Wrong. Most people think that Economics is only related to money, investments, or stuff like that, but Economics is really a way of thinking about everything. In order to explain this “risk vs. rewards” trade-off, I’ll use an example that you might find yourself more identified with: Imagine you want to ask the girl you have a crush on for a date. There is a risk, she might directly say no and make you feel stupid or worthless, not getting a reward and also imposing a psychological cost. But she might also say yes, giving you the reward of going out with her, having a good time, and maybe even start a relationship. You always have the possibility of not taking the risk and do nothing about it. Again, there is no right or wrong answer here… Well, this case is more particular. The classic model of rational choices, what I explained at the beginning of this article, would say that for some bets, the rational decision is to take them. Let’s say you are given a bet with a probability of 90% of getting 100 dollars and a probability of 10% of losing 50 dollars. It is easy to see that (for a risk-neutral person), the expected value of this bet is E[100, -50; 0.9, 0.1] = 85 dollars, as it is positive, you should take it. The model doesn’t end there, it also takes into consideration your degree of risk-aversion. Let’s think of the case of being extremely risk-averse. Think of the 50 dollars you could lose as the last $50 dollars on your bank account. Even if the bet is “safe” and a “good” one, you won’t be willing to risk those last $50 because it’s all you got… Or maybe because of that exact reason, you are very willing to risk them for a positive large expected gain. This decision solely relies on your degree of risk-aversion, which is determined by many personal characteristics. So if you are a guy that is willing to take risks, you will probably ask the girl out. How else can this affect you? Let’s go back to the decision of choosing your major (and for the purpose of this explanation we will only consider the salary part of the decision). There might be a major, let’s say Petroleum Engineering, that you know that will give you a really high salary with no doubt, but maybe you don’t like at all. Perhaps your passion is painting, and you want to go to Art School. There is a chance that you will become a great painter and get a lot of money by selling your paintings in hipster galleries in Chelsea, New York, but there is also a chance that you won’t. So your decision depends on how much risk you are willing to take. Doing something you love with the risk of a low salary, or doing something you are not interested in with a guaranteed high salary? Again (for the third time), there is no right or wrong decision. But my point doesn’t end here. How reliable is this risk-aversion model? According to it, a person who turns down a bet with 50/50 chance of losing $1000 and winning $1050 would also turn down a bet with 50/50 chance of losing $20000 or winning any sum of money, yes, any sum of money. We know that risk-aversion and expected-utility make sense, but do the theoretical results hold in real life? For those of you who want to know a little bit more about this, check out this paper by Matthew Rabin.
Now think of any decision you had to make. You can always analyze it from the perspective of these three different trade-offs. And, as I have shown you in every case, the rational models of decisions don’t always hold. This doesn’t end here, there are many other “anomalies” that are extremely interesting. For example, people tend to be more affected from losses than from gains. The psychological cost of losing ten dollars can be perceived as larger than the psychological benefit of winning ten dollars. There is another important bias called the “endowment effect”. Individuals give goods a greater valuation when they physically own them. For example, I could give a shirt I see at a store the value of $20, however, after I buy it, it is no longer just a shirt, it is MY shirt, even if it is the same as any other one in the store, I feel differently about it now, so my valuation for it is larger. This phenomenon has quite an important effect in different markets. Think about the housing market. You want to sell your house, the one you have been living in for the past 20 years. It is very likely that the value you give to it (hence, the price you are asking for it), is higher than the real value of the house in the market, which could end up in a number of sales in the market that is under the optimal one. What are the implications of this? Think of real-state agencies. If instead of selling the house by yourself, a third-party is designated with that duty, they don’t have any personal attachment to the house, so the way they valuate the asset might be different (A more detailed explanation of these biases can be found in this paper). What about peer-effects? Maybe the rational model says that you would rather stay at home on a Friday night and work on your essay for Monday, but what if all your friends are going to a party and you feel the peer-pressure of joining them? Decision-making is not individual, but more of a collective process.
Finally, another funny thing to have in mind is the power of ignorance when it comes to decision-making. Let’s focus on the stock market for a second. The Herzfeld Caribbean Basin Fund Inc. is a stock part of NASDAQ that has the ticker symbol (abbreviation used to uniquely identify publicly traded shares) of CUBA, yes, like the country. Needless to say, the company has absolutely no relationship to Cuba whatsoever, apart from the abbreviation of their name. However, when President Obama announced that the US and Cuba would restore full diplomatic ties, the stock price of The Herzfeld Caribbean Basin Fund Inc. skyrocketed… What? But didn’t we say that there was no relationship between the two? Like my mother used to say: “La ignorancia es atrevida”. Check out the graph showing the closing price and volume of the stock:
Okay, let’s accept it, we all make mistakes, so it is understandable that the moment after the restore of diplomatic relationships was announced, some of us might have immediately thought of buying the stock abbreviated as CUBA. But… Nowadays information can reach any corner of the planet in no time. You can open your Twitter feed and check what is going on second to second. Stock brokers are known to be one of the groups of people that are up to date with everything because a delay of one second in information can cost them foregoing millions of dollars. So, how can we explain this?
I want to finish this article with a practical example of decision-making. Apart from behavioral economics, my main field of interest is development economics and impact evaluation. Growing up in one of the poorest cities of a “third-world” country has made me to be always interested in poverty alleviation. In my opinion, a key determinant of poverty is Education, maybe it is the most important one, considering all the externalities related to it. But it has been proven that most poor people underinvest in education, so what explains this? I will present you a very simple model of schooling-decision, that sums up the factors that determine a poor household’s decision of sending or not their children to school, and how policymaking can affect this:
Consider and individual who lives for 2 periods: t ∈ {1, 2}. He has an instantaneous utility in each period: V. The individual discounts his future utility by β. Finally, he chooses to study in the first period or not. If he decides to study in the first period, he has to incur the cost of schooling (Direct costs: transportation, tuition, uniforms, etc. Opportunity costs: effort, less leisure, forgone earnings, etc). This cost of schooling is represented by C. However, by attending school, he can get a return B in period 2 with probability p and get nothing with 1-p.
If he doesn’t go to school: EU[S=0] = V + βV
If he decides to go to school: EU[S=1] = V-C+β[V+pB]
So he will go to school if EU[S=1] > EU[S=0] → βpB > C
In words, his decision depends on how much he discounts the future, the probability of getting a benefit, the size of that benefit, and the cost of schooling. This result by itself doesn’t say much, but you can add variations in the variables to get more interesting results, and even add things like Credit and the positive externalities of schooling. So why does this matter? To see what can policymakers do to increase educational attainment for poor households. For example: reducing cost of schooling with subsidies, reduce credit constraints so individuals can make optimal investments, increase the probability that benefits to schooling are realized, use behavioral tools to address the time-discounting characteristics of the poor. There is a very interesting paper that deals with this, a good mix between behavioral and development economics, feel free to check it out here.
Decision-making is very complex. Every decision you face involves several types of trade-offs. From deciding between now or later to deciding between personal and common benefits and between risk or rewards. The beauty of decision-making (yes, for the fourth and last time) is that there is no wrong decision. There might be decisions that go against the classical economic models of rationality, in fact, most of the decisions in real life, but that doesn’t mean they are wrong, and it doesn’t mean the models are wrong either. It just means that the models are a way of simplifying reality, which is not always precise, but is helpful to understand how people make decisions. Decision-making is where psychology, economics, sociology, math, and several other fields, collide. Even if we have spent years trying to understand the logic behind it, there is still a lot to do.
