Apples and Oranges: A Critique of Utilitarianism — Part I
This is part I of a 4-part series on utilitarianism. This material is related to my forthcoming book, WILD PROBLEMS: A GUIDE TO THE DECISIONS THAT DEFINE US, coming on August 9. In that book, I critique what I call narrow utilitarianism — the day-to-day calculus of pleasure and pain — as a basis for making big life decisions. In the process of writing the book, I realized that utilitarianism is embedded in various ways in how economists think about decision-making at both the societal and individual level. I also realized that the utilitarian approach tends to encourage economists to focus on the measurable at the expense of what is not measurable and that this can lead us badly astray, a theme that is at the heart of my book. So this essay is in some sense the intellectual backstory of the issues that I write about in WILD PROBLEMS as well as expanding the ideas in the book beyond the focus on individual decision-making and looking at the implications for public policy.
“If your knowledge cannot be measured it is meager and unsatisfactory”
— Lord Kelvin
Let’s give Lord Kelvin his due. Without measurement, progress is impossible. Consider a farmer in early times, long before the advent of modern agriculture. How is he doing? Probably not so well. Hovering on the edge of subsistence, barely able to keep his family and himself fed winter to winter, he is desperate to find ways to improve his annual crop of apples. No matter how he’s doing, he wants to do better. To do better, he must measure. He must count. How many apples did he harvest this year vs. last year? Is his crop going up or going down? Arithmetic is born.
He notices that in places where animals graze, the trees seem to do better compared to where animals do not graze. So he tries to fertilize part of his land with animal manure. Is this a good idea? To find out, he must count the number of apples he gets this year vs last year from the areas he has fertilized. If he is a very wise farmer he will realize that any change in the crop may be due to the something other than just the fertilizer, like a change in rainfall. If he is especially wise, he will compare this year’s crop to the average of the last five years to try to take out the impact of variations in other factors. Statistics is born.
Perhaps this farmer has gone beyond self-sufficiency. He’s not only able to feed himself and his family, but he’s sufficiently productive to grow more apples than his family can eat. He takes his extra apples into town and sells them, using the money to buy other things he cannot make easily for himself. Commerce is born.
Once he interacts with others through buying and selling, he can ask a deeper question. Maybe it’s a mistake to grow apples. By growing apples, he gives up the chance to grow something else. And now economics is born, the idea of trade-offs and opportunity cost. To really know how he’s doing, the farmer needs to know something of what is possible, how he might be doing if he made different choices.
The farmer decides to reduce the size of his apple orchard and replace some of his apple trees with orange trees. He doesn’t know — yet — that orange trees won’t grow well in climates where apple trees grow well. But he plunges ahead and he wonders whether he made a mistake. Does his meager orange harvest make up for the apples he’s no longer growing? To answer the question, he needs to compare apples and oranges.
I apologize for this elaborate joke — comparing apples and oranges is of course, the metaphor for two things that aren’t comparable. The metaphor captures why we have an urge to quantify, to measure. A farmer must implicitly or explicitly compare apples and oranges if she wants to know how he’s doing.
Before arithmetic and accounting, the farmer’s measure of well-being would be loose, vague, and indeterminate. He would have a crude idea of how he was doing by how often his children complained about being hungry. He could look out into the field and eyeball the amount of land covered with apple or orange trees. He could have a feel for how long it takes to harvest different fruits. He might understand something about the inherent variability of some crops relative to others
A lot, a few, some, abundant, cornucopia, meager, decent, pretty good, amazing are all imprecise ways to capture how the farmer is doing. But the desire for certainty and a desire to improve, pushes the farmer toward more precise measures of her harvest. And once she has more than one crop, the problem becomes particularly challenging. Saying you have a lot of apples but not so many oranges isn’t that helpful. A farmer needs a way to compare apples and oranges.
This is especially urgent as the number of crops and the different kinds of livestock get more numerous. In any one year, the harvest of some crops will be higher and some lower. Some crops may be big and unblemished while others are small and barely edible. He may have a growing herd of cattle but because he planted less corn, his chickens may have less to eat and may not be laying as many eggs as before.
How can the farmer wrap his head around this complexity?
Some answers are more helpful than others.
Counting the number of apples, oranges, ears of corn, eggs, chickens or even to go crazy — the number of grains of barley — and so on and adding them up, does give you a measure of success but it’s not an actionable measure because there is no inherent value to such a count — it correlates poorly with anything the farmer actually cares about. Another obvious measure, weight, has the same problem.
A better measure, once there is a market for his produce and livestock, is the dollar value of everything he produces. Multiply the number of ears of corn by the price per ear, the numbers of eggs times the price of an egg and so on. The result is his income. Accounting is born.
The dollar value of each crop is the common denominator that lets you compare apples and oranges. It’s a very good measure because the farmer’s income is crucially correlated with not starving to death, correlated with being able to rise above abject poverty and correlated with being able to hire someone to fix the leaky roof of the farmhouse or send the kids off to school instead of having them work on the farm.
Imagine a chart that has all of the different aspects of the farmer’s output for a year. The chart would not just have the total number of eggs but the number of eggs graded by size and quality. It wouldn’t have just tons of corn, but how many tons were good enough for human consumption and how many were only fit to feed to the cattle. It’s overwhelming.
Income takes all the complexity of the various crops and their myriad characteristics and converts that complexity into a single number. It’s really magic.
The chart would have many different rows and columns and you can always add more rows and columns to capture this or that characteristic of the harvest. And many of the entries in that chart — a subjective measure of the color of the apples from this year to the next, for example — would not be comparable in any way. As life gets more complicated, it gets harder and harder to eyeball the chart from one year to the next and figure out whether you’ve made progress. Measuring income takes the loose, vague, and indeterminate information embodied in the chart and turns it into something precise, accurate, and indispensable — a single number.
If that isn’t magic, what is?
This urge to quantify, to transform information that is loose, vague, and indeterminate into something more precise such as a single number is hard to resist. It answers our desire for certainty and it opens the door to progress. It’s a fantastically powerful tool for management. It has two glorious features. First, it’s simple. In contrast to the chart, you can see it instantly.
There’s a second important feature of a single number that captures the success or failure of this year’s harvest. You can compare it to last year’s number. You can see if it’s bigger or smaller. You can project next year’s number and when next year arrives, you can see if your plans came to fruition.
When you consider buying a house you look at its location and how many bedrooms it has and the size of the kitchen and so on. But every house has irregularities and different shapes. So we usually rely on a single number — the square footage — to figure out which one is bigger. I may care independently about the size of the kitchen because I love to cook (or don’t care about cooking at all), but the square footage sure beats a list of the different rooms and their respective sizes. The ability to boil complexity down to a single number so I can make comparisons is very powerful.
The mathematical name for numbers that describe physical concepts like area is scalar. The origin of the word is the Latin word for ladder, or scala — something that helps you to climb. It’s the same Latin word for scale as in the things that help you measure or the verb as in to scale the highest peaks — to rise.
Scalars make it easy to put things on a single scale, to make them comparable. They simplify complicated things. We are really good as humans at heavier, higher, taller, shorter, bigger, smaller. We are really good at comparing numbers and deciding whether one is bigger, smaller, or the same as the other. 1000 is bigger than 10. 17.3 is bigger than 17.1. Making these comparisons are so easy we never think about it.
It also opens the door to mistakes. In “This is Spinal Tap,” guitarist Nigel Tufnel played by Christopher Guest shows Marty DiBergi played by Rob Reiner the special amplifiers that his band uses and points to the knobs that control volume:
Nigel Tufnel: The numbers all go to eleven. Look, right across the board, eleven, eleven, eleven and…
Marty DiBergi: Oh, I see. And most amps go up to ten?
Nigel Tufnel: Exactly.
Marty DiBergi: Does that mean it’s louder? Is it any louder?
Nigel Tufnel: Well, it’s one louder, isn’t it? It’s not ten. You see, most blokes, you know, will be playing at ten. You’re on ten here, all the way up, all the way up, all the way up, you’re on ten on your guitar. Where can you go from there? Where?
Marty DiBergi: I don’t know.
Nigel Tufnel: Nowhere. Exactly. What we do is, if we need that extra push over the cliff, you know what we do?
Marty DiBergi: Put it up to eleven.
Nigel Tufnel: Eleven. Exactly. One louder.
Marty DiBergi: Why don’t you just make ten louder and make ten be the top number and make that a little louder?
Nigel Tufnel: [pause] These go to eleven.
Q.E.D! Christopher Guest looks at Rob Reiner like he’s a moron. A five-year old knows that 11 is bigger than 10. But not all elevens are created equal.
This kind of mistake isn’t the typical one. Go back to the farmer. It’s hard to remember that the single number, income, is not a perfect measure. It’s hard to remember that it’s not the only thing we care about. It’s hard to remember that it’s only an approximation of how we’re doing even if all you care about is material well-being.
Some economists claim you don’t need economic models to understand the world. You just need data. Let the numbers do the talking. But numbers are mute. They only speak with our help. And what we say they are saying, inevitably involves a model of the world. But the model is veiled by the simplicity of the numbers.
In today’s Washington Post, [https://www.washingtonpost.com/local/dc-hunger-report/2020/10/01/1770590c-0337-11eb-8879-7663b816bfa5_story.html] as I write these words, there is a story with the headline:
District’s food insecurity rate estimated to be 16 percent, up from 10.6 percent before pandemic
I don’t know what food insecurity is. The article does not explain it. Presumably it is related to hunger. That it has increased by such a large amount is disturbing but it’s hard to know just how disturbing without knowing what exactly is meant by food insecurity and more importantly, how the measure of 16 percent was determined. I suspect most people read the story and concluded, as the story suggests we should, that food insecurity is up a little over 50% in the nation’s capital. Definitely alarming.
The story does link to a report from the Office of Planning in the Mayor’s office of Washington, DC. I click through and open the report. The first paragraph mentions:
the critical importance of ensuring that every resident in the District of Columbia has access to healthy, affordable, and culturally appropriate food.
Is that the definition of food insecurity? Not quite. That is explained in the second paragraph which reads in its entirety:
“Food insecurity” is a term defined by the U.S. Department of Agriculture that refers to a lack of consistent access to enough food for an active, healthy life.
Food insecurity is definitely a bad thing. Or at least it could be depending on how it is defined. But as this makes clear, it’s a concept that is loose, vague, and indeterminate. Food insecurity has some connection to hunger and malnutrition but measuring it is inevitably a subjective exercise to quantify a subjective concept and give it the patina of an objective measure. When we see a comparison between 16% and 10.6% our brains see something that looks objective.
And is there anything more objective than the decimal point? When last year’s food insecurity measure for Washington DC was calculated, it wasn’t rounded up to 11 or made vague by saying “a little more than 10%.” It was 10.6%. Most jokes about economists are cruel. Perhaps the cruelest is this one:
Q: How do you know macroeconomists have a sense of humor?
A: They use decimal points.
Meaning that the macroeconomics forecaster isn’t content to predict that unemployment will be higher or lower next year or next quarter or next month relative to the present. The forecaster will usually predict the precise number carried out to at least one decimal point to give the forecast the air of precision that sometimes comes with scientific measurement. Decimal points, to quote W.S. Gilbert of Gilbert and Sullivan, “give artistic verisimilitude to an otherwise bald and unconvincing narrative.”
And it turns out, when you dig deeper into the report on food insecurity, the 16 percent number is a projection. The report says that food insecurity is projected to be at least 16 percent. I want to suggest that when you see the headline, your mind decides that there is a reliable objective fact that food insecurity is at least 50% worse than it was before. But that’s not an objective fact. What it really means requires some digging and even some digging may not be enough.
The desire for certainty, the desire to turn a matrix into a scalar, a chart of attributes into a single number, is not just understandable, it’s often the only way to make progress on a problem. If food insecurity is indeed on the rise and policies are put in place to help reduce food insecurity, measurement gives you some chance to know if the policies helped or hurt.
Without measurement, as Lord Kelvin points out, we can’t know if our knowledge is reliable. That’s the upside of measurement. The downside is that inevitably, almost any measurement we use embodies a set of assumptions that are easily forgotten.
I’m about 66 inches from the bottom of my feet to the highest point on my head. I say “about” because precise measurement is difficult. It depends on how I stand and the challenge of measuring anything with perfect precision. But that I am 5’6” is reasonably described as a fact, partly because any deviation from that precision is quite small relative to the correct, unobserved, not easily ascertained exact height I am right this minute on October 2, 2020. It’s also reasonably described as a fact because you as the consumer of that fact have probably used a tape measure or a ruler or yardstick and are very aware of getting an accurate measure of height within say 1/64th of an inch.
The same would be true of the temperature right now in your house or outside. The weather app on my phone says it is 63 degrees Fahrenheit in my house. That is surely not the exact temperature just outside my front door or even 30 feet away from it. There’s no decimal point and the precise temperature will vary whether it is measured over asphalt in the street or in the yard where the grass reacts differently to sunlight. But like height, when I tell you the temperature is 63 that’s close enough for deciding whether to wear a sweater, short-sleeves, or a heavy parka if you are planning on leaving the house.
But is the 16% food insecurity rate for Washington DC a fact? It looks like a fact because it’s a number. Our brains associate numbers with what we call facts, numbers that have standards of measurement that we use or have used often, like measuring height or observing the temperature on a thermometer. But many things that look like objective facts are what we might call scientism — things that have the appearance of science but without the reliability of science. Our brain struggles to keep them apart.
I don’t know whether it is a good idea or a bad idea to try to quantify a subjective concept like hunger, poor access to healthy food, or poor access to affordable food. What I am saying here is that your brain struggles to remember that a number like 16% as a measure of food insecurity is not a simple fact like height or temperature. But your brain, if it’s like mine, is prone to treating it like a simple fact.
The practice of turning a matrix into a scalar is all around us. It’s the air we breathe, the intellectual water we swim in. It’s the human response to information overload. It takes a complex array of reality and turns it into something we can wrap our heads around.
Which movie directed by Rob Reiner is better, “This is Spinal Tap” which IMDB users rate at 7.9 or “When Harry Met Sally” which IMDB users rate at a mere 7.6 as of October 28, 2020? Obviously TISP is a better movie than WHMS, right? Of course not. Better is meaningless here. The higher rating for TISP means one thing and one thing only — the people who rated TISP on average gave it a higher rating than WHMS. That does not tell you what you might really want to know — if you can only watch one tonight, which one will you enjoy more? It certainly doesn’t mean that TISP is a better measure in any objective sense.
And I bring this up mainly because This is Spinal Tap is the only movie at IMDB that is rated out of 11. For better or worse, 10 is still the maximum rating you can use. But the average is listed as 7.9 out of 11. I cannot decide if IMDB should have let users give Spinal Tap an eleven.
Let’s return to the farmer. The farmer’s income last year is much closer to an objective measure than the level of food insecurity in Washington, DC. And income is a particularly good scalar because it aligns so nicely with something the farmer cares deeply about — achieving a level of material well-being that could mean the difference between life and death. So measuring the total value of the harvest seems like the perfect way to measure success and solve the apple and oranges problem. Now the farmer can track his performance and know whether he is doing better this year than last year. The concept of income takes a messy complex matrix and converts into a clean, simple scalar that allows the farmer to evaluate progress. But a farmer will still find it challenging to use wisely.
Maybe the farmer starts to get excited when he sees that the value of this year’s harvest exceeds that of last year. He starts to wonder if he can make it even bigger next year. In his ambition, he might forget that some crops have harvests that are highly variable. The dollar value of this year’s harvest doesn’t capture any of the uncertainty surrounding next year. The orange harvest is very sensitive to changes in temperature. The farmer may expand his orange crop in hopes of making more money and forget the downside risk that comes from that decision.
Or in his excitement to expand the size of the harvest, he may forget that devoting more time to making money comes at a cost. That means less time for his family, which isn’t counted in the dollar value of the harvest. Or he may damage his health working longer hours with less sleep. These intangibles are hard to keep in mind. Like the person who looks for the car keys under the streetlight, the farmer may forget what else there is to care about.
What gets measured gets managed. But I’m saying something stronger here. If we are not careful, what gets measured is all we manage. We don’t just pay more attention to what is in the light. We forget what is in the shadows. We forget about the rest of the things that do not get captured in measures we become accustomed to studying and using.
To be continued. In Part 2, I go deeper into how our love of scalars can lead us astray…