What Is The Point Of Studying Economics
It’s not to get a job
“Why study economics?” is related to, but more interesting than the more commonly answered question “Is economics a science?”. At most, the second question should just be a stepping stone to the first, which would let us ask questions about economics as something specifically taught. This is, I think, what Jacob Viner meant when saying “Economics is what economists do.”. If studying economics is to be able to do what economists do and become a kind of scientist, then that is a sufficient reason to study it.
Getting a degree in economics isn’t the only way to study economics. A better phrasing therefore might be, outside of any particular institutional context,, “Why get a handle on price theory?”. Most of what matters in an undergrad econ degree - for instance, getting to see how statistics are used in social sciences - rarely connect with the kind of econ 101 type price theory we’re all familiar with from the news and internet arguments. Contrariwise, one of the best ways to learn economics - to lose a bunch of arguments with people who know what they are talking about! - isn’t going to get you a degree at the end.
CVAR Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
So, however we study economics, we don’t really know what that means or what the payoff is yet. The problem is that economics is not a science with a complex but unitary deductive structure, in the way rigid body mechanics is. Instead, economics may be a discipline whose practitioners are scientists, in the way that mechanical engineering is. Familiarity with a particular toolbox for engaging real-world engineering situations rather than a total model of the world viewed from an economic lens. So, who are the scientists?
Science And Scientists
This might take us a little far afield, but it is helpful to think through where the term “scientist” came from in the first place. Had things gone a little differently, we might have been in a position today to earnestly call them “nature-pokers” (from the German ‘natur-forscher’).
“Formerly, the ‘learned’ embraced in their wide grasp all the branches of the tree of knowledge… But these days are past; … we adopt the maxim ‘one science only can one genius fit’. The mathematician turns away from the chemist; the chemist from the naturalist; the mathematician, left to himself, divides himself into a pure mathematician and a mixed mathematician who soon part company. And thus science, even mere physical science, loses all traces of unity. A curious illustration of this result may be observed in the want of any name by which we can designate the students of the knowledge of the material world collectively. … ‘Philosophers‘ was felt to be too wide and to lofty a term … ‘savans’ was rather assuming, besides being French instead of English; some ingenious gentleman proposed that, by analogy with ‘artist’, they might form ‘scientist’, and added that there could be no scruple in making free with this termination when we have such words as sciolist, economist, and atheist--but this was not generally palatable;…”
William Whewell, Review of “On The Connexion Of The Physical Science”
Under this definition, a ‘Scientist’ is a person with a narrow but deep expertise of particular practices centered on some specific science, which in turn can be defined as a collection of sentence complexes which are judged by being right or wrong by participants in the scientific endeavor. Keynes used the term ‘group’ to describe these sentence complexes, but that has since been definitively taken by us mathematicians.
This is an abstract, but surprisingly robust definition of scientific practice that gets us away from the “realism vs anti-realism” debates that produced so much hot air in the 90s. Whatever you want to call them, typical examples of these kinds of scientific sentence complexes can be found in the early chapters of George B Airy’s book Gravitation, which deals with the motion of planets (in the old sense of all celestial objects except the earth and fixed stars). Despite having no equations, the book is quantitative and mathematically sophisticated.
Anyway, one sentence complex of the book concludes “The motion of the planet subject to such forces [as in the restricted three body problem] would be nearly the same as if it was revolving in an elliptic orbit, and this orbit was the same time revolving around its focus, turning in the same direction as the planet goes round and always carrying [the planet] on [the ellipse’s] circumference. And this is the easiest way of representing to the mind the general effect of this motion; the physical cause in such explanations as that above.”.
This sentence complex is (a special case of) Newton’s revolving orbits theorem, which is used to understand the orbit of the Earth’s moon. For a treatment in modern notation, see Chapter 10 of Chandrasekar’s Newton’s Principia For The Common Reader. This particular sentence complex as variously used by Newton, Airy and Chandrasekar is a particular kind of scientific sentence complex, a so-called ‘scientific law’. There are four features that identify this as a scientific law:
Each noun is related to each other in multiple ways.
If the forces due to gravity of the various planets were not separately measurable from the revolving orbit, then the sentence could only serve as a definition of those particular forces. This would make it a tautology, not a scientific law.
The sentence is universally quantified.
Though intended to understand the moon’s motion, nothing about the moon is specifically mentioned. Rather it should describe all possible orbits in a class of situations that hopefully contains the moon’s orbits. The moon here would be just one example of that law being followed.
It is approximate.
All scientific laws contain approximations in two ways. First, scientific laws reduce the space of relations between the variables to only the relevant relations. Second, there is the definition of which particular situations the universal quantification above applies to. This will always include edge cases, which are filled in by advances in technique (Kuhn’s Normal Science)
It is ahistorical.
It is important to remember – from a pedagogical as well as philosophical perspective – that scientific laws are not the only kind of scientific sentence complex found in sciences. There are natural histories, descriptions of measurements, recipes, existence statements, purely statistical predictions, tautologies and perhaps dozens of others.
Furthermore, even much of the best and most idealized scientific discourse is not a member of these scientific sentence complexes, but is about the complexes. This includes statements about the truth, falsity, complexity, etc.. of the sentence complexes themselves. In certain ways, scientific explanation is about laying out rigorous low-level sentence complexes that can then be abstracted from to create meta-complexes that actually have some verbal, narrative, or technical explanatory power for some question people are interested in. This is why philosophers around the turn of the 20th century saw such promise in a “scientific society,” whether they were American Pragmatists or partisans of Red Vienna.
Let’s look at these distinctions in the context of the recent Diamond & Dyvbig Nobel prize. Their partisans claim that Diamond & Dyvbig created a scientific law which governs the effect of financial crises on economic output, in contrast to all previous examinations, which are figured by these partisans as mere natural histories. Others claim that this account is pretentious and hyperbolic. At best on this view Diamond & Dyvbig are responsible for developing an existence proof within a particular kind of model which shows that a behavior well-known and well-described in natural history can be expressed in terms of the model.
These kinds of debate are common and ordinary in science, even if they are seen as threatening the very legitimacy of “economics” as a whole in some quarters. The reason that the partisan above is so proud to call Diamond & Dyvbig a “scientific law” is that scientific laws are often placed at the core of the explanation of the advancement of science. This centering has a rhetorical aspect (which has been much remarked upon), but it also has a practical aspect. The four aspects of scientific laws listed above provide a sufficient explanation why it would matter if D-D were a “law”: scientific laws are often particularly simple kinds of sentence complexes. This simplicity allows them to be most easily taught in undergraduate education without the broad explanatory background provided by “natural history,” which is often pedagogically left until after a student has decided they actually care about the phenomenon in question.
Now we are back at “Why study economics?” with a new point of view: if one can quickly learn scientific laws related to the economy, then one can easily build the skills of an economist and scientist.
But all this simplification only raises a harder question “Are there actually any scientific laws relating to the economy at all?”.
Economic Laws And Economic History
In Economic Laws And Economic History, Charlie Kindleberger acutely examined four economic scientific laws and the extent to which they explain actual economic history. Since it is Kindleberger, we don’t have to worry about the economic scientific laws being silly things like “demand curves are always downward sloping” or anything like that. We can focus on the better question of: do these easily taught portions of the economics syllabus actually do more work than the more complicated parts? Expressed in terms of elasticities, the four laws he examines are
The income elasticity of physical input is between zero and one.
Unskilled labor has a near perfectly elastic supply at the subsistence wage.
If the cross elasticity of two goods are high (such as goods which are qualitatively the same but differ in quantitatively), then the quantity of one good in a particular market will tend to zero
Every good has a very high elasticity of substitution with itself.
Kindleberger shows that these laws – together with the underlying Marshallian system - form a powerful point of view for thinking about, in order: technological growth, economic growth, economic instability and international trade. Take Engel’s law, which implies relative spending on necessities is inversely proportional to the level of output. Couldn’t one try to use it in reverse: suppress consumption to increase growth? This is no idle question, ask Khruschev.
One can see therefore how these economic scientific laws provide a very good reason to study the geometric system of economics that started with Marshall. By examining a few special cases - infinite elasticity of substitution, perfect elasticity of supply or positive subunity income elasticity - one can derive very reasonable economic scientific laws by the principle of continuity.
But this is very far from the Econ 101ism that everyone is familiar with. The main problem in using these “laws” is, as Kindleberger argues, the mistaking of these market laws for causal principles. These laws are ‘merely’ the comparative statics (relative market positions) of exogenous (unexplained within the model) growth. Another way of thinking about it might be to say that accounting laws provide one set of boundary conditions for a monetary using economy, while the four laws above provide an additional not-strictly-collinear set of boundary conditions. By applying both, we can constrain the range of questions about which economics can reason. By using those constraints intelligently, it is possible to develop model closures that can explain actually-interesting economic questions.
Engel’s law, for instance, is a powerful tool for reasoning about growth as long as growth is occurring. During the stagnation of the late 70s, the Soviet system of consumption repression came under extreme pressure when the forces creating the growth it was meant to multiply were no longer present. The law didn’t break down, but the economy moved out of the space well-explained through the combination of accounting and economic constraints that had been established at the time. The space the economy moves through is history. History as a space is more like the space of systems with (irreversible) phase changes.
There is another reason to be wary of the lack of causal mechanism in Econ 101 style price theory: when these laws’ range of applicability are suspended, they rarely have “characteristic suspensions” whereby suspension is a result of the action of some determinant alternative force. This means an explanation by scientific law may be specious without giving an indication why it is specious, frustrating Deborah Mayo’s severe tests.
Stigler’s Theory Of Price begins by looking at a typical situation in which the law of one price is applied: washing machines. Washing machines have a high but finite elasticity of substitution with themselves: there is limited but definite price dispersion. Stigler seems to link this price dispersion with the search cost of price discovery. Thus there is only one apparent lesson: the only way to reduce price dispersion is to get the consumer better information.
But Stigler has taught no such lesson. There are many alternate explanations of price dispersion. First of all, the basic Marshallian premise that the buyer of a good can usually sell at the same price applies to the wholesale but not consumer market of washing machines. That is, warehouse dealers may be willing to make a market for wholesalers, but not consumers. Second, firms might target reducing the elasticity of substitution of their goods via, for instance, branding (monopolistic competition, in econ lingo). Third, Marshall’s assumptions of well-informed dealers may be violated by the lack of sophistication of small business dealers of washing machines. Finally, washing machines may have highly price inelastic demands and multiple user classes (firms that generate a lot of laundry - hotels, hospitals, etc - vs households), in which case the marginal conditions underlying elasticity as a concept are unstable (as pointed out by Ken Boulding in his review of Paul Samuelson’s Foundations).
The problem isn’t so much flaws in the “scientific laws” or the impossibility of “scientific laws” in economics; the problem is the narrow range of applicability even when “laws” are described adequately. If the economy can quickly move out of the range of states within which these laws are applicable, it demonstrates more forcefully that economic “laws” can never replace natural history. Yet in science, this is not even a discussion. No one seriously argues that further development of theory no longer requires experimentalism.
Now, we can sort of answer our original question and see why it was necessary. The study of economics begins with economic scientific laws simply because they are simple. These supplement knowledge of actual historical conditions and social forces. However, economic scientific laws cannot substitute for that knowledge. The social and technical role that “forecasting” plays here is simply too enormous and complicated to get into, because then you need oodles of agents and goals and things.
Yet this leaves us firmly in the company of Joan Robinson: “The purpose of studying economics is not to acquire a set of ready-made answers to economic questions, but to learn how to avoid being deceived by economists.”.
In other words, when ideally applied, economic scientific laws can explain why someone got hired but not why it was you. What could be more aufklarung than that?
CVAR Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.